{"id":632,"date":"2026-07-26T13:22:53","date_gmt":"2026-07-26T13:22:53","guid":{"rendered":"https:\/\/matlani.in\/?p=632"},"modified":"2026-07-26T13:22:53","modified_gmt":"2026-07-26T13:22:53","slug":"ani-media-pvt-ltd-v-open-ai-opco-llc","status":"publish","type":"post","link":"https:\/\/matlani.in\/?p=632","title":{"rendered":"ANI Media Pvt. Ltd. v Open AI OPCO LLC"},"content":{"rendered":"\n<div data-wp-interactive=\"core\/file\" class=\"wp-block-file\"><object data-wp-bind--hidden=\"!state.hasPdfPreview\" hidden class=\"wp-block-file__embed\" data=\"https:\/\/matlani.in\/wp-content\/uploads\/2026\/07\/ANI-vs-Open-AI.pdf\" type=\"application\/pdf\" style=\"width:100%;height:600px\" aria-label=\"Embed of ANI vs Open AI.\"><\/object><a id=\"wp-block-file--media-b0747c12-73e8-43df-8f7b-2f3c8832543a\" href=\"https:\/\/matlani.in\/wp-content\/uploads\/2026\/07\/ANI-vs-Open-AI.pdf\">ANI vs Open AI<\/a><a href=\"https:\/\/matlani.in\/wp-content\/uploads\/2026\/07\/ANI-vs-Open-AI.pdf\" class=\"wp-block-file__button wp-element-button\" download aria-describedby=\"wp-block-file--media-b0747c12-73e8-43df-8f7b-2f3c8832543a\">Download<\/a><\/div>\n\n\n\n<p>Markdown File:<\/p>\n\n\n\n\n\n\n\nCS(COMM) 1028\/2024     Page 1 of 135\n\n## * IN  THE  HIGH  COURT  OF  DELHI  AT  NEW  DELHI\n%        Judgment reserved on: 27\nth\n## March, 2026\nJudgment pronounced on: 24\nth\n## July, 2026\n\n## I.A. 45300\/2024\n## IN\n## +  CS(COMM) 1028\/2024, I.A. 45301\/2024 &#038; I.A.26192\/2025\n\nANI MEDIA PVT. LTD.      &#8230;..Plaintiff\n## Through: Mr. Sidhant Kumar, Mr. Akshit Mago,\n## Ms.  Manyaa  Chandok,  Ms.  Anshika\nSaxena and Ms. Lahar Jain, Advocates\n\n## Versus\n\nOPEN AI OPCO LLC              &#8230;..Defendant\nThrough: Mr. Amit Sibal, Senior Advocate with\n## Mr.  Sanjeev  Kapoor,  Mr.  Nirupam\n## Lodha,   Mr.   Madhav   Khosla,   Mr.\n## Gautam Wadhwa, Ms. Moha Paranjpe,\n## Mr.  Abhi  Udai  Singh  Gautam,  Ms.\n## Rebecca  Cardoso,  Mr.  Hardik  Malik,\n## Ms.  Vanshika  Thapliyal,  Mr.  Rajat\n## Bector, Advocates\nMr. Akhil Sibal, Senior Advocate with\n## Mr. Aditya Gupta,  Ms.  Asavari  Jain,\n## Mr. Shuvam Bhattacharya,  Ms.  Vani\n## Kaushik,   Ms. Riddhie Bajaj,   Ms.\n## Jahnavi Siddhu, Ms. Aishwarya Kane\nand Mr. Sauhard Alung, Advocates for\nIGAP\/Intervenor in I.A. 4616\/2025\nMr. Kapil Sibal, Senior Advocate and\n## Mr. Arvind P. Datar, Senior Advocate\nwith Mr. Shashank Mishra, Ms. Akshi\nRastogi,  Mr.  Parv  Kaushik  and  Mr.\nSuvaroop  Saha  Roy,  Advocates  for\n## Broadband India Forum\n\n\n\nCS(COMM) 1028\/2024     Page 2 of 135\n\n(BIF)\/Intervenor in I.A. 10766\/2025\n## Ms.  Haripriya  Padmanabhan,  Senior\nAdvocate     with     Mr.     Shrutanjaya\n## Bhardwaj,  Mr.  Akshat  Agrawal,  Mr.\n## Tushar  Srivastava,  Mr.  Shourya  Das\n## Gupta,   Ms.   Siddhi   Nagwekar,   Ms.\nYashi  Bajpai  and  Mr.  Yash  Tayal,\nAdvocates     for     Flux     AI     Labs\/\nIntervenor in I.A. 4615\/2025\n## Mr.  Ankit  Sahni,  Ms.  Kritika  Sahni,\n## Mr.   Chirag   Ahluwalia,   Mr.   Mohit\n## Maru, Ms. Tanisha Sharma,\nAdvocates for the Federation of Indian\nPublishers, Intervenor in I.A. No. 470\nof 2025\n## Mr. Rajshekhar Rao, Senior Advocate\nand   Mr.   Chander   M.   Lall,   Senior\nAdvocate  with Mr.  Ameet  Datta,  Mr.\n## Harsh Kaushik, Ms. Riddima Sharma,\n## Mr. Akshay Nagarajan, Ms. Rishikaa,\nMs.  Gauri  Khanna and  Ms.  Annanya\nMehan, Advocates for\nDNPA\/Intervenors  in I.A.  2199\/2025\n&#038; IMI\/Intervenors in I.A. 4027\/2025\nMr. Adarsh Ramanujan with\n## Mr.  Parth  Singh,  Advocate, Amicus\n## Curiae\n## Professor Arul George Scaria, Amicus\n## Curiae.\n\n## CORAM:\n## HON\u2019BLE MR. JUSTICE AMIT BANSAL\n## JUDGMENT\n## AMIT BANSAL, J.\nI.A. 45300\/2024 (under Order XXXIX Rules 1 and 2 CPC)\n\n\n\n\nCS(COMM) 1028\/2024     Page 3 of 135\n\n## INDEX\nPREFACE &#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;- 4\nWORKING OF LLMS &#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;- 8\nA. TRAINING DATA &#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8211; 8\nB. DATA PREPARATION AND MODEL TRAINING &#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212; 9\nC. POST-TRAINING &#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;- 10\nD. USER INTERACTION &#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8211; 10\nE. RETRIEVAL-AUGMENTED GENERATION (RAG) &#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8211; 11\nPROCEEDINGS IN THE SUIT&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212; 11\n## ISSUE NO.4\n## WHETHER THE COURTS IN INDIA HAVE JURISDICTION TO ENTERTAIN THE PRESENT LAW\n## SUIT CONSIDERING THAT THE SERVERS OF THE DEFENDANTS ARE LOCATED IN THE\nUNITED STATES OF AMERICA &#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;- 13\nSUBMISSIONS &#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;- 13\nDISCUSSION AND ANALYSIS &#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8211; 19\ni. Whether this court has territorial jurisdiction to entertain the present suit. &#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8211; 19\nii. Whether the Copyright Act, 1957 would apply in relation to training claim since according to\nOpen AI training takes place on servers located outside India &#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8211; 20\nFINDINGS&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;- 22\n## ISSUE NO.2\n## WHETHER THE USE BY THE DEFENDANTS OF PLAINTIFF\u2019S COPYRIGHTED DATA IN\n## ORDER TO GENERATE RESPONSES FOR ITS USERS, WOULD AMOUNT TO INFRINGEMENT\nOF THE PLAINTIFF\u2019S COPYRIGHT. &#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212; 23\nSUBMISSIONS &#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;- 23\nDISCUSSION AND ANALYSIS &#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8211; 34\ni. Whether Open AI memorizes and regurgitates ANI\u2019s copyrighted literary works in the form of\nresponses? &#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;- 37\nii. Whether ChatGPT\u2019s responses are substantial reproduction of ANI\u2019s copyrighted literary\nworks? &#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212; 42\nFINDINGS&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;- 60\n## ISSUE NO.1\n## WHETHER THE STORAGE BY THE DEFENDANTS OF PLAINTIFF\u2019S DATA (WHICH IS IN THE\n## NATURE OF NEWS AND IS CLAIMED TO BE PROTECTED UNDER THE COPYRIGHT ACT,\n## 1957) FOR TRAINING ITS SOFTWARE I.E., CHATGPT, WOULD AMOUNT TO INFRINGEMENT\nOF PLAINTIFF\u2019S COPYRIGHT &#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8211; 61\nSUBMISSIONS &#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;- 61\n## ISSUE NO.3\n## WHETHER THE DEFENDANTS\u2019 USE OF PLAINTIFF\u2019S COPYRIGHTED DATA QUALIFIES AS\n\u2018FAIR USE\u2019  IN TERMS OF SECTION 52 OF THE COPYRIGHT ACT, 1957. &#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212; 68\nSUBMISSIONS &#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;- 68\nDISCUSSION AND ANALYSIS OF ISSUE NOS. 1 AND 3 &#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;- 86\n\n\n\nCS(COMM) 1028\/2024     Page 4 of 135\n\nA. Scope of Section 14(a)(i) of the Copyright Act, 1957 &#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8211; 86\nB. Legislative History of Section 52(1)(a)&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;- 92\nC. Interpretation of Section 52(1)(a) of the Copyright Act &#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8211; 93\nD. Scope of Section 52(1)(a) &#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;- 95\nI. PURPOSE TEST &#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;- 96\na. Whether commercial use is entitled to protection under Section 52(1)(a)(i) &#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8211; 97\nb. Requirement of \u201cnon-infringing copy\u201d under Section 52(1)(a) of the Copyright Act. &#8212;&#8212;&#8212;&#8211;104\nc. Whether use by Open AI amounts to \u201cprivate or personal use, including research\u201d &#8212;&#8212;&#8212;&#8212;-108\nII. FAIRNESS TEST &#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8211; 115\na. Whether Open AI\u2019s use of ANI\u2019s original literary works is limited to training its LLMs\nunderlying ChatGPT?&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8211;123\nb. Whether the usage of ANI\u2019s literary works by Open AI would result in economic competition\nand would prejudice the legitimate interests of ANI, thereby causing actual or potential damage\nto ANI? &#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8211;124\nc. Whether the functions performed by Open AI through ChatGPT  serve the overall public\ninterest? &#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;-128\nFINDINGS&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212; 129\n\nBALANCE OF CONVENIENCE AND IRREPARABLE INJURY &#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212; 129\nSUBMISSIONS &#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212; 129\nDISCUSSION AND ANALYSIS &#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;- 132\nFINDINGS&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212; 134\nCONCLUSION &#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;- 134\n\n\n\n\n\n\n\nCS(COMM) 1028\/2024     Page 5 of 135\n\n## PREFACE\n&#8211; Innovations  affecting  the  works  protected  under  the  Copyright  Act,\n1957 (hereinafter the \u2018Copyright Act\u2019), namely literary, dramatic, musical and\nartistic works, cinematograph films, and sound recordings, have consistently\ngiven  rise  to interpretive  challenges in  the  application  of  traditional  legal\nprovisions.\n&#8211; While  tracing  the  history  of copyright  law,  it  is  noteworthy  that\nglobally, copyright law and technological innovation have long existed in a\ndynamic relationship, traceable at least to the advent of the Gutenberg Press\nin the fifteenth century, well before the enactment of the Statute of Anne in\n## 1710\n## 1\n## .\n&#8211; With time, successive technologies have evolved, enabling new forms\nof  creative  expression  and  widening  avenues  for  dissemination,  disrupting\nsettled copyright arrangements. Some of these technologies worth referring to\nare player pianos, radio broadcasting, cable television, photocopying, home\nvideo  cassette  recorders,  digital  downloading  and  streaming,  and,  more\nrecently,  artificial  intelligence  (hereinafter \u2018AI\u2019).  Therefore,  technological\nadvancement continues to generate novel questions that necessitate periodic\nre-examination of established copyright doctrine.\n&#8211; It  is  the  judiciary,  as  the  primary  forum  for  adjudication,  that  is\ninvariably  called  upon  to  address  the  nuanced  legal  questions  arising  from\nsuch  disruptive  technologies,  often  resulting  in  conflicts,  as  in  the  present\n\n## 1\nIn 1709 British parliament enacted the Statute of Anne; short title: Copyright Act 1709 8 Anne c.21; long\ntitle: An Act for the Encouragement of Learning, by vesting the Copies of Printed Books in the Authors or\npurchasers of such Copies, during the Times therein mentioned. Named after Anne, Queen of Great Britain,\nthis was the first copyright statute in the Kingdom of Great Britain, and the first full-fledged copyright statute\nin the world. It was enacted in the regnal year 1709 to 1710, and entered into force on April 10, 1710.\n\n\n\nCS(COMM) 1028\/2024     Page 6 of 135\n\ncase.  Therefore,  the Courts  are  required  to  step  in  to  fill  the  gap  between\nadvancing technologies and existing laws.\n&#8211; AI  is  currently  a  technology  of  widespread  interest,  and  India  is\nconsidered  a  forerunner  in  this  field. In  response  to international  AI\ncompetition, Sarvam AI of India launched its first domestic Large Language\nModel  (hereinafter  \u2018LLM\u2019) on   18\nth\n## February   2026\n## 2\n. Research   and\ndevelopment of similar LLMs is also underway in India to make India an AI\nhub. Data is the oil for LLMs to work efficiently.\n&#8211; Procuring publicly available data by AI innovators for training LLMs\nwithout  the  permission  of  the  respective  copyright  holders  potentially  will\nraise legal issues, as arising in the present suit. The present suit has been filed\nby the plaintiff (hereinafter \u2018ANI\u2019) against the defendant (hereinafter \u2018Open\nAI\u2019) alleging unauthorised use of ANI\u2019s copyrighted works by Open AI. ANI\nclaims copyright infringement on two counts:\ni. Training claim &#8211; copying and storage of ANI\u2019s data to train Open AI\u2019s\nLLM; and\nii. Reproduction  or  Output  claim &#8211; reproduction of ANI\u2019s works in the\noutput produced by Open AI\u2019s LLM through its application ChatGPT.\n&#8211; This case is a perfect illustration of how the law must continually catch\nup  with  technology.  The  case  has  to  be  adjudicated  on  the  anvil  of  the\nCopyright Act, originally enacted in 1957, though amended from time to time.\nClearly, when the Copyright Act or any of its amendments were enacted, the\nlegislature  could  not  have  imagined  the  vast  technological  advances  in  the\n\n## 2\nAroon  Deep, Bengaluru  firm  unveils  two  AI  language  models,  The  Hindu  (Feb.  18,  2026),\nhttps:\/\/www.thehindu.com\/sci-tech\/technology\/bengaluru-firm-unveils-two-ai-language-\nmodels\/article70648438.ece. , accessed on 20\nth\n## February 2026.\n\n\n\nCS(COMM) 1028\/2024     Page 7 of 135\n\nfield of AI.\n&#8211; In  these  uncharted  areas, Courts  often  take  aid  from  judgments  of\nforeign Courts  on  similar  issues.  On  various  occasions,  while  dealing  with\nIntellectual Property laws, the Indian Courts have recognized the importance\nof   global   doctrinal   interdependence.   The Courts   laid   emphasis   on\nharmonization  of  basic  principles  of  laws  in  foreign  jurisprudence  with  the\nnational  laws,  especially  when  the  national  laws  do  not  indicate  a  contrary\nintent\n## 3\n.  However,  while  doing  so,  the Courts  must  be mindful that  the\nunderlying legislation in foreign jurisdictions may be quite different from that\nprevailing in India.\n&#8211; For the purposes of adjudication of the present application for interim\ninjunction filed by ANI, this Court had framed issues on 19\nth\n## November, 2024,\nwhich are set out below:\n\u201cI. Whether  the  storage  by  the  defendants  of plaintiff\u2019s data (which is in\nthe nature of news and is claimed to be protected under the Copyright\nAct,  1957)  for  training  its  software  i.e.,  ChatGPT,  would  amount  to\ninfringement of plaintiff\u2019s copyright.\nII. Whether  the  use  by  the  defendants  of plaintiff\u2019s copyrighted data in\norder to generate responses for its users, would amount to infringement\nof the plaintiff\u2019s copyright.\nIII. Whether the defendants\u2019 use of plaintiff\u2019s copyrighted data qualifies as\n\u2018fair use\u2019 in terms of Section 52 of the Copyright Act, 1957.\nIV. Whether the Courts in India have jurisdiction to entertain the present\nlawsuit considering that the servers of the defendants are located in the\nUnited States of America.\u201d\n\nEven though Issue no. 3 refers to \u2018fair use\u2019 [as it was framed on the first date\nof hearing] it should be read as \u201cfair dealing\u201d which is the expression used in\nSection 52(1)(a) of the Copyright Act.\n\n## 3\nIntex Technologies v. Telefonaktiebolaget LM Ericsson, 2023:DHC:2243-DB.\n\n\n\nCS(COMM) 1028\/2024     Page 8 of 135\n\n&#8211; Since the question of this Court\u2019s jurisdiction to adjudicate the present\napplication is foundational to the determination of the remaining issues, the\nCourt shall  first  consider  Issue  No.  4.  Thereafter,  in  view  of  its  broader\nbearing on Issue Nos. 1 and 3, the Court shall examine Issue No. 2. As the\nAmici Curiae and counsel for the parties were ad idem that Issue Nos. 1 and\n3  are  intertwined,  particularly  in  light  of  Open AI\u2019s  plea  of  fair  dealing  in\nresponse  to  the  allegations  relating  to  storage, the  said issues  shall  be\nconsidered together, subsequently.\n&#8211;  With   the   aforesaid   backdrop,   I   proceed   to   decide   the   present\napplication.\n## WORKING OF LLMS\n&#8211; It is necessary to understand the functioning of LLMs for the purpose\nof  adjudicating  the  present  application,  as  the  issues  identified  herein  arise\nfrom,  and  are  intrinsically  connected  with,  the  manner  in  which  LLMs\noperate. Accordingly, it is apposite to first examine the technical functioning\nof LLMs.\n&#8211; Based on the submissions of the parties and learned Amici Curiae and\nalso from the common general knowledge available, the steps involved in the\nworking of LLMs, may be summarised in the following paragraphs.\n## A. Training Data\n&#8211; A large language model is a computational system that can predict the\nnext word from previous words. That is, given a context or prefix of words, a\nlanguage  model  assigns  a  probability  distribution  over  the  possible  next\nwords.\n&#8211; For  enabling  a  model  to  predict  subsequent  words,  the  model  is  first\n\n\n\nCS(COMM) 1028\/2024     Page 9 of 135\n\ntrained on an enormous text corpus with datasets of hundreds of billions of\nwords, generally scraped from the web, which is referred to as the \u2018raw data\u2019.\n&#8211; The \u2018raw  data\u2019 used for training LLMs would include automatically\nscraped  data  from  publicly  available  sources  on  the  internet,  as  well  as\nlicensed   materials.   The   web   text   is   usually   taken   from   corpora   of\nautomatically  crawled  web  pages,  such  as  the  Common  Crawl.  The  data\ncomprises materials, including but not limited to books, articles, blog posts,\netc., so that, based on the aforesaid data, the model can make predictions.\n&#8211; Pre-training data collected from the web is filtered for both quality and\nsafety.  Quality  filtering  uses  automated  tools  to  score  documents  based  on\ntheir usefulness and reliability, often by comparing them with trusted sources\nsuch  as  books  and  reputable websites,  while  avoiding  content  containing\npersonal  information  or  adult  material.  This  generally  improves  language\nmodel performance. Safety filtering aims to remove harmful or toxic content,\ntypically using existing toxicity-detection tools.\n## 4\n\n&#8211; Website owners can indicate that they do not want their sites crawled\nby web crawlers.\n## 5\n\nB. Data preparation and Model training\n&#8211; The raw data is divided into smaller pieces called tokens, each assigned\na \u2018Token ID\u2019, which is just a number. This is called tokenisation. This process\nis the first step in training most of the LLMs.\n## 6\n\n\n## 4\nMatthew Sag and Peter K. Yu, \u2018The Globalization of Copyright Exceptions for AI Training\u2019 (Emory Law\nJournal, 2025, Forthcoming) https:\/\/papers.ssrn.com\/sol3\/papers.cfm?abstract_id=4976393.\n## 5\nDaniel Jurafsky and James H. Martin. 2026. Speech and Language Processing: An Introduction to Natural\nLanguage  Processing,  Computational  Linguistics,  and  Speech  Recognition  with  Language  Models,  3rd\nedition. Online manuscript released January 6, 2026. https:\/\/web.stanford.edu\/~jurafsky\/slp3.\n## 6\nDaniel  J.  Gervais  et  al.,  \u2018The  Heart  of  the  Matter:  Copyright,  AI  Training,  and  LLMs\u2019  (2024)\nhttps:\/\/papers.ssrn.com\/sol3\/papers.cfm?abstract_id=4963711.\n\n\n\nCS(COMM) 1028\/2024     Page 10 of 135\n\n&#8211; Thereafter,  each  token  is  converted  to  a  vector via an \u2018embedding\u2019\nprocess. \u2018Vectors\u2019 are  a  numerical  representation  of  semantic  information\nabout a \u2018token\u2019.\n&#8211; In the training exercise, the LLM is repeatedly fed parts of the training\ndata where a word or a sequence of words is masked, and it is made to predict\nthe next words or sentences, which are then compared with the original \u2018raw\ntext\u2019 to verify correctness. The initial predictions may not be accurate, but with\neach repetition, the model readjusts and refines its prediction to reach the most\nprobable statistical output in each scenario.\n&#8211; The  choice  of which  word to generate in LLMs is done  by sampling\nfrom the distribution of possible next words.\n## C. Post-training\n&#8211; Thereafter,  the  model  is  again  fine-tuned  to  reduce  prediction  errors\nand  to  follow  instructions  better,  for  example,  to  answer  questions,  give\nsummaries, write code, translate sentences, and so on. It does this by being\ntrained on a special corpus containing many texts with both instructions and\ntheir correct responses.\n&#8211; The model is then adjusted to give more useful responses and to avoid\nproducing harmful or offensive content.\nD. User interaction\n&#8211; A user instructs LLM via a prompt, a text string that a user issues to a\nlanguage  model  to  get  the  model  to  do  something  useful  by  iteratively\ngenerating tokens conditioned on the prompt. The process of finding effective\nprompts for a task is known as prompt engineering.\n## 7\n\n\n## 7\nMark Lemley, \u2018How Generative AI Turns Copyright Law Upside Down\u2019 (2024) 25(2) Sci. &#038; Tech. LR 36.\n\n\n\nCS(COMM) 1028\/2024     Page 11 of 135\n\nE. Retrieval-Augmented Generation (RAG)\n&#8211; Retrieval-Augmented Generation (\u2018RAG\u2019) is an innovative feature of\nLLMs that optimises their output. This feature references an authoritative\nknowledge  base  outside  of  its  training  data  sources  before  generating  a\nresponse. This feature does not predominantly rely on training data to generate\noutput. Instead, it uses an information retrieval component that utilises the\nuser prompt to pull information from external data storage.\n## 8\nThe LLM uses\nboth training data and external information to generate better responses via\n## RAG.\n&#8211; For the sake of convenience, \u2018LLM(s)\u2019 and \u2018ChatGPT\u2019 has been used\ninterchangeably in the judgment and must be read likewise. When reference\nis made to \u2018ChatGPT search\u2019 function, it has been specifically mentioned in\nthe judgment as such.\n## PROCEEDINGS IN THE SUIT\n&#8211; The  present  suit,  accompanied  by  an  application  seeking  interim\ninjunction and other ancillary applications, was first listed before this Court\non 19\nth\nNovember 2024. On the same date, four issues were framed and the\nCourt appointed two Amici Curiae to assist the Court.\n&#8211; Thereafter, multiple intervention applications were filed in the present\nsuit, the details of which are given below:\n29.1. I.A. 470\/2025 has been filed on behalf of Federation of Indian\nPublishers,  a  national  representative  body  of  the  Indian\npublishing industry.\n\n## 8\nWhat is RAG? &#8211; Retrieval-Augmented Generation AI Explained &#8211; AWS, https:\/\/aws.amazon.com\/what-\nis\/retrieval-augmented-generation\/.\n\n\n\nCS(COMM) 1028\/2024     Page 12 of 135\n\n29.2. I.A.  2199\/2025 has  been  filed  on  behalf  of Digital  News\nPublishers Association and its members, which are digital news\nplatforms.\n29.3. I.A. 4027\/2025 has been  filed  on  behalf  of  the Indian Music\nIndustry, which is an organization established to defend, preserve\nand develop the rights of the Phonogram Producers.\n29.4. I.A. 4615\/2025 has been filed on behalf of Flux Labs AI Private\nLtd.,  a  Generative  Artificial  Intelligence  enabled  start-up\ncompany.\n29.5. I.A. 4616\/2025 has been filed on behalf of IGAP Project LLP, a\nthink  tank  focusing  on  the  intersection  of  law,  policy,  and\nemerging technologies, including AI.\n29.6. I.A. 10766\/2025 has been filed on behalf of Broadband India\nForum, an independent policy forum and think tank.\n&#8211; The intervenors in I.A. 470\/2025, I.A. 2199\/2025 and I.A. 4027\/2025\nare broadly supporting the case of ANI and the intervenors in I.A. 4615\/2025,\nI.A.  4616\/2025  and  I.A.  10766\/2025  are  broadly  supporting  the  case  of\nOpen AI.\n## 31. On  4\nth\nJuly  2025,  this Court allowed  the  aforesaid  intervention\napplications and permitted the intervenors to make submissions limited to the\nlegal issues framed on 19\nth\n## November 2024.\n&#8211; ANI, Open  AI as  well  as  the  intervenors  have  filed  their  written\nsubmissions along with judgments in support.\n&#8211; Submissions on behalf of the parties were heard on 21\nst\n## February 2025,\n## 10\nth\n## March 2025, 18\nth\n## March 2025, 28\nth\n## March 2025, 2\nnd\n## April 2025, 22\nnd\n## April\n## 2025, 29\nth\n## April 2025, 4\nth\n## July 2025, 23\nrd\n## July 2025, 30\nth\n## July 2025, 5\nth\n## August\n\n\n\nCS(COMM) 1028\/2024     Page 13 of 135\n\n## 2025,  18\nth\n## August  2025,  12\nth\n## September  2025,  23\nrd\n## September  2025,  17\nth\n\n## October 2025, 7\nth\n## November 2025, 21\nst\n## November 2025, 23\nrd\n## January 2026,\n## 20\nth\n## February 2026, 20\nth\nMarch 2026 and 27\nth\nMarch 2026, when the judgment\nwas  reserved  and  the  parties  were  given  liberty  to  file  their  closing\nsubmissions.\n&#8211; ANI, Open AI and the two Amici Curiae have filed their concluding\nwritten submissions.\n34.1 The concluding written submissions dated 2\nnd\nApril 2026 sent on behalf\nof Professor Arul George Scaria via email to the Court Master, have also been\ntaken on record.\n&#8211; With the aforesaid backdrop, I proceed with my issues-wise analysis.\nIssue No.4 Whether  the Courts  in  India  have  jurisdiction  to\nentertain the present law suit considering that the servers\nof the defendants are located in the United States of\n## America\n\n## SUBMISSIONS\nSubmissions on behalf of the Defendant (OPEN AI)\n## 9\n\n&#8211; Mr. Amit Sibal, senior counsel appearing on behalf of Open AI made\nthe following submissions:\n36.1. ANI has joined two separate causes of action in the present suit i.e. &#8211; (i)\nuse of publicly available data by Open AI on servers located outside India to\ntrain LLMs underlying ChatGPT (\u2018training claim\u2019) and, (ii) the responses\nreceived by ANI from ChatGPT allegedly similar to content over which ANI\nholds  copyright  (\u2018reproduction  claim\u2019).  Therefore,  ANI  has  to  establish\n\n## 9\nSince the issue of jurisdiction has been raised by the defendant, it is deemed appropriate that the submissions\nof the defendant are recorded first.\n\n\n\nCS(COMM) 1028\/2024     Page 14 of 135\n\njurisdiction qua both causes of action.\n36.2. Insofar as the training claim is concerned, it is an undisputed position\nthat the training takes place outside India on the basis of data that is stored in\nservers  outside  India.  Therefore,  merely  because  the  jurisdiction  can  be\nestablished in respect of the reproduction claim, this Court would not have\njurisdiction in respect of training claim.\n36.3. The  Copyright Act  and  in  particular  Section  62  thereof  read  with\nSection  1(2)  of  the  Copyright  Act,  confers  jurisdiction  only  within  the\nterritory of India. The Copyright Act was not intended to have extra-territorial\njurisdiction.\n36.4. The  reliance  placed  by  Professor  Scaria  on  the  judgment  of a\nCoordinate Bench in Neetu Singh v. Telegram\n## 10\nis misplaced insofar as the\njurisdiction in respect of training claim is concerned. In Neetu Singh (supra),\nthe infringement occurred within the jurisdiction of this Court and in that\ncontext, the Court observed that it would be immaterial if telegram servers are\nlocated  outside  India.  Similarly,  reliance placed on  the  judgment of  the\nSupreme Court in Exphar SA v. Eupharma Laboratories\n## 11\nand the judgment\nof Division Bench in Blueberry Books v. Google India\n## 12\nis misplaced.\n36.5. The test laid down by the judgment of the Division Bench in Banyan\nTree Holding v. A. Murali Krishna Reddy\n## 13\nfor the purposes of jurisdiction\nis limited. The purposeful availment test as laid down in Banyan Tree (supra)\ncannot be used to confer jurisdiction in respect of causes of action occurring\noutside India.\n\n## 10\n2022 SCC OnLine Del 2637.\n## 11\n## (2004) 3 SCC 688.\n## 12\n2016 SCC Online Del 3338.\n## 13\n2009 SCC OnLine Del 3780.\n\n\n\nCS(COMM) 1028\/2024     Page 15 of 135\n\n36.6. Insofar as the reproduction claim is concerned, at the time of using the\nChatGPT  service,  the  user  agrees  to  \u2018terms  of  use\u2019  of  ChatGPT,  which\nprovides jurisdiction of the Courts in San Francisco, California.\n\nSubmissions on behalf of the Plaintiff (ANI)\n&#8211; Mr.  Sidhant  Kumar,  counsel  appearing  on  behalf  of ANI  made  the\nfollowing submissions:\n37.1. In terms of Section 62(2) of the Copyright Act, copyright owners can\nfile suit for infringement in the jurisdiction where they reside or conduct\nbusiness.  Therefore,  the  physical  location  of  Open AI\u2019s  servers  is  of  no\nconsequence. ANI\u2019s  principal  and  registered  office  is  located  within  the\njurisdiction of this Court.\n37.2. Even in terms of Section 20(b) of the Code of Civil Procedure, 1908\n(hereinafter  \u2018CPC\u2019),  since  Open  AI  operates  an  interactive  website  that\nenables  commercial  transactions  within  the  jurisdiction  of  this Court, i.e.\nOpen AI collects subscription fee from users located in this jurisdiction, Open\nAI would be amenable to jurisdiction of this Court.\n37.3. Open AI actively attracts the consumers through advertising and public\noutreach  efforts  in  India,  that  are  within  the  jurisdiction  of  this Court.\nPurposeful  engagement  of  ANI\u2019s  consumers  and  a  targeted  marketing\nsufficiently demonstrate compliance of Section 20(c) of the CPC.\n37.4. The various infringement activities alleged in the plaint have taken\nplace within the jurisdiction of this Court. The disputes raised in the present\nsuit are independent of \u2018terms of use\u2019 of ChatGPT. The acts that constitute\ncopyright  infringement  are  independent  of  any  contractual  agreements\nbetween the users and Open AI. The statutory remedies such as those provided\n\n\n\nCS(COMM) 1028\/2024     Page 16 of 135\n\nunder Section 62 of the Copyright Act cannot be waived through contractual\nstipulations.\n\nIntervenors supporting ANI\ni. Submissions on behalf of the Intervenor \u2013 DNPA (I.A. 2199\/2025)\n&#8211; Mr. Rajshekhar Rao, senior counsel appearing on behalf of intervenors\nin I.A. 2199\/2025 made the following submissions:\n38.1. The copies\/reproduction of Indian works which are first published in\nIndia and thereafter made available by Open AI to attract Indian users, would\nimply  that  at  the  very  least  part  of  the  cause  of  action  arises  within  the\njurisdiction of this Court. Therefore, in terms of Section 62 of the Copyright\nAct, read with judgment of the Supreme Court in Indian Performing Rights\nSociety v. Sanjay Dalia\n## 14\n, this Court would have the jurisdiction to entertain\nthe present suit.\n38.2. Open AI is entering into commercial transactions within the jurisdiction\nof this Court and specifically targeting the customers. Reliance is placed on\n## Banyan Tree (supra).\n38.3. Neither the \u2018terms of use\u2019 of Open AI\u2019s ChatGPT, nor the fact that the\nservers located outside India, would oust the jurisdiction of this Court under\nSection 62 of the Copyright Act as well as Section 20 of the CPC.\n\nii. Submissions on behalf of the Intervenor &#8211; Indian Music Industry (I.A.\n## 4027\/2025)\n&#8211; Mr. C.M. Lall, senior counsel appearing on behalf of intervenors in I.A.\n4027\/2025 made the following submissions:\n\n## 14\n## (2015) 10 SCC 161.\n\n\n\nCS(COMM) 1028\/2024     Page 17 of 135\n\n39.1. Jurisdiction  is  also  made  out  on  the  parameters  of \u2018purposeful\navailment\u2019 as laid down in Banyan Tree (supra). In the present case, availment\ncan be demonstrated since Open AI has \u2018paid subscribers\u2019 of its services\nwithin the jurisdiction of this Court and the said subscribers and users interact\nwith the service within the jurisdiction.\n\niii. Written Submissions filed by Mr. Ankit Sahni, Advocate on behalf of the\nIntervenor \u2013 Federation of Indian Publishers (FIP) (I.A. 470\/2025)\n&#8211; This Court would have jurisdiction under Section 62 of the Copyright\nAct. In this regard, reliance has been placed on the judgment of the Supreme\nCourt in Exphar\n## 15\n(supra) and Dabur India Ltd. v. K.R. Industries\n## 16\nto submit\nthat  Section  62  of  the  Copyright Act  was  inserted  by  the  Parliament  to\nprescribe  an  additional  ground  for  attracting  jurisdiction  over  and  above\nSection 20 of CPC.\n40.1 Since Open AI\u2019s services are accessible within the jurisdiction of this\nCourt and the outputs generated by Open AI\u2019s ChatGPT application infringes\nthe copyright of ANI, a substantial part of cause of action arises within the\njurisdiction of this Court to invoke Section 20 of CPC.\n\nSubmissions of Mr. Adarsh Ramanujan, Amicus Curiae\n&#8211; Mr. Ramanujan made the following submissions:\n41.1. The location of Open AI\u2019s servers outside India is irrelevant as ANI has\ninvoked Section 62 of the Copyright Act. There is no challenge by Open AI\nto  the  assertion  of ANI  that  its  principal  place  of  business  is  within  the\njurisdiction of this Court.\n\n## 15\n## \u00b6 13.\n## 16\n## (2008) 10 SCC 595, \u00b6 32.\n\n\n\nCS(COMM) 1028\/2024     Page 18 of 135\n\n41.2. Insofar as the submission of Open AI is concerned that the infringement\nrelating to the training claim occurs outside the jurisdiction of this Court, what\nhas to be seen is that the training process involves access of data from India\nand transmission of the said data from India. Therefore, it cannot be said that\njust because the training occurs outside the jurisdiction of this Court, this\nCourt would not have jurisdiction. The training claim has to be seen in a\nholistic manner which would necessarily involve access and transmission of\ndata from India before the same is stored overseas.\n\nSubmissions of Professor Arul George Scaria, Amicus Curiae\n&#8211; Prof. Scaria made the following submissions:\n42.1. There is no dispute raised to the fact that ANI has its principal office in\nNew Delhi. Therefore, this Court would have jurisdiction under Section 62(2)\nof the Copyright Act.\n42.2. Since Open AI is making its active services available to the users in\nIndia, including those in Delhi, the Courts in India would have jurisdiction in\nterms of Section 20 of the CPC.\n42.3. Reliance is placed on the judgment of the Supreme Court in Exphar\n(supra)  and  the  judgment  passed  by  the  Division  Bench  of  this Court in\nBlueberry Books (supra) and the judgment of a Coordinate Bench in Neetu\n## Singh (supra).\n42.4. Insofar  as  aspect  of  the  territoriality  of  Indian  Copyright  Act  is\nconcerned, he supports the contention of Open AI that since the entire training\nprocess occurs outside India merely because Open AI accesses data from India\nwould not make the infringement subject to Indian Copyright Act.\n\n\n\nCS(COMM) 1028\/2024     Page 19 of 135\n\n42.5. In Getty Images v. Stability AI\n## 17\nthe plaintiff in the United Kingdom\n(\u2018UK\u2019) did not press the training claim as the training took place outside UK.\n\n## DISCUSSION AND ANALYSIS\n&#8211; As discussed earlier, since the issue of jurisdiction goes to the root of\nthe matter, it would be taken for consideration at the beginning.\n&#8211; On the aspect of jurisdiction, based on the objections raised by Open\nAI, essentially the following two issues arise for consideration:\ni. Whether this Court has territorial jurisdiction to entertain the present\nsuit.\nii. Whether the Copyright Act would apply in relation to training claim\nsince  according  to  Open AI  training  takes  place on  servers located\noutside India.\n\ni. Whether this Court has territorial jurisdiction to entertain the present\nsuit.\n&#8211; Insofar as the territorial jurisdiction of this Court to entertain the present\nsuit is concerned, it is an admitted position that ANI has its principal place of\nbusiness as well as its registered office within the jurisdiction of this Court.\nTherefore, in terms of Section 62(2) of the Copyright Act, this Court would\nhave the territorial jurisdiction to entertain the present suit.\n&#8211; It is also an admitted position that Open AI specifically targets\/offers\nits services to users\/subscribers across India, including those located within\nthe jurisdiction of this Court.\n&#8211; Even alleged infringing activities mentioned in the plaint, i.e. responses\n\n## 17\n[2025] EWHC 38 (Ch).\n\n\n\nCS(COMM) 1028\/2024     Page 20 of 135\n\ngenerated by Open AI in India based on the prompts by ANI, took place within\nthe jurisdiction of this Court.\n&#8211; Accordingly, at this prima facie stage, I hold that this Court would have\nthe territorial jurisdiction to entertain the present suit in terms of Section 20\nof the CPC as well as Section 62(2) of the Copyright Act.\n\nii. Whether the Copyright Act would apply in relation to training claim\nsince according to Open AI training takes place on servers located\noutside India\n&#8211; It has been argued on behalf of Open AI that its LLMs are trained\noutside India and the training data is stored and processed in the servers\nlocated in the United States of America (hereinafter \u2018US\u2019). On behalf of Open\nAI, it is contended that the Indian Copyright Act would not apply outside the\nterritory  of  India.  In  support  of  its  contention,  reliance  is  placed  on  the\njudgment of the Division Bench in Blueberry Books (supra). In the said case,\nAmazon, which was a defendant in the suit took the defence that the infringing\nact took place exclusively in the US. Though Amazon admitted that it has\nbusiness\/commercial activities in India, the alleged infringing activity did not\noccur within the jurisdiction of this Court.\n&#8211; The Division Bench, in the said case, upheld the territorial jurisdiction\nof this Court. Insofar as the argument of Amazon with regard to the infringing\nactivity taking place in the US was concerned, the Division Bench did not\nuphold the contention of Amazon. Paragraph 27 of Blueberry Books (supra)\nis set out below:\n\n\u201c27.  As  far  as  inapplicability  of  the  provisions  of  the  Act  to\ninfringements  within  the  USA  are  concerned,  we  note  that  the\nsubmission is based on Amazon\u2019s arguments to that effect. Its unilateral\n\n\n\nCS(COMM) 1028\/2024     Page 21 of 135\n\nassertion  that  there  are  downloading  restrictions,  based  on  a  bare\nreading of the documents &#8211; or that only 33 downloads took place, for a\nsmall amount, are not relevant; they are contentious and accepting them\nwould be giving disproportionate weight to the pleadings of one party\nand precluding the rights of the plaintiff, finally.\u201d\n\n&#8211; In my view, the storing of ANI\u2019s works on the US servers is a terminal\nstep in the chain of events which begin from access of copyrighted works\nfrom India and transmission of the same abroad. There can be no doubt that\nCopyright  Act  cannot  have  jurisdiction  outside  the  territory  of  India.\nHowever, the Copyright Act does not require this Court to sever the chain of\nevents and examine only the last step. On a prima facie view, Open AI\u2019s\nargument is untenable as it would lead to evasion of Indian copyright law by\ninfringers who shift the terminal link to servers abroad.\n&#8211; In Neetu Singh (supra), a Coordinate Bench of this Court has observed\nthat merely because Telegram chooses to locate its servers in Singapore, that\nwould not affect the right of the copyright owners, to sue Telegram in India.\nThe relevant observations from the aforesaid judgment are set out below:\n\u201c46. In view of the above factual and legal position, in the opinion of\nthis Court, merely because Telegram chooses to locate its server in\nSingapore,  the  same  cannot  result  in  the  Plaintiffs\u2019 &#8211; who  are\ncopyright  owners  of  course  materials &#8211; being  left  completely\nremediless against the actual infringers, especially in order to claim\ndamages and avail of other legal remedies in accordance with law. If\nsuch  an  argument  is  accepted,  in  the  current world  where  most\ndissemination  happens  through  online  messaging  services  and\nplatforms, IP violations would go completely unchecked. This cannot\nbe the intention of law. The provisions of the IT Act and the Rules\nmade therein have to be construed harmoniously with the rights and\nremedies provided to the copyright owners under the Copyright Act.\nIndian Courts are competent to decide issues relating to infringement\nof copyright and the mere fact that Telegram is operating a messaging\nservice in India which chooses not to locate its servers in India cannot\n\n\n\nCS(COMM) 1028\/2024     Page 22 of 135\n\ndivest the Indian Courts from dealing with copyright disputes or divest\ncopyright owners from availing their remedies in Indian Courts. In\nthe  present  age  of  cloud  computing  and  diminishing  national\nboundaries in data storage, conventional concepts of territoriality\ncannot be strictly applied. The dynamic evolution of law is essential\nto ensure appropriate remedies in case of violation of copyright and\nother IP laws.\u201d\n[Emphasis supplied]\n\n&#8211; The aforesaid reasoning would be fully applicable in the present case.\nIn Neetu Singh (supra), even though the data was stored in servers outside\nIndia, the same was being circulated within the jurisdiction of this Court.\n&#8211; In  my  view,  the  act  of  Open AI  in  scraping  of ANI\u2019s  copyrighted\nmaterial and training of its LLM cannot be entirely divorced from the output\nclaim. Since the output claim is based on training of the LLM, and the output\nis  reproduced  within  the  jurisdiction of  this Court,  I  am  not  inclined  to\nseparate the two causes of action.\n## FINDINGS\n&#8211;  On a prima facie view, it cannot be said at this stage that this Court\ndoes not have the territorial jurisdiction to entertain the present suit or that the\nalleged infringement on account of training would involve extra-territorial\napplication of the Indian Copyright Act.\n&#8211; As  discussed  earlier,  next,  I  would  take  up  Issue  No.  2  for\nconsideration.\n\n\n\nCS(COMM) 1028\/2024     Page 23 of 135\n\nIssue No.2 Whether  the  use  by  the  defendants  of  Plaintiff\u2019s\ncopyrighted data in order to generate responses for its\nusers, would amount to infringement of the Plaintiff\u2019s\ncopyright.\n## SUBMISSIONS\nSubmissions on behalf of the Plaintiff (ANI)\n&#8211; Mr.  Sidhant  Kumar,  counsel  appearing  on  behalf  of  ANI  made  the\nfollowing submissions:\n57.1. LLMs  generate  responses  by  matching  user  input  with  the  vector\ndatabase  and  predicting  the  token  sequences.  This  is  the  process  of  \u2018de-\ntokenization\u2019. Once a response is formed, detokenization converts tokens into\nreadable text by reassembling words, phrases, and sentences. Detokenization\nreconstructs raw data into its original textual form, converting training data\n(including ANI\u2019s works) back into a format readable by humans. Therefore,\nthe whole process amounts to reproduction of raw data.\n57.2. ANI has provided instances of ChatGPT\u2019s unauthorized reproduction\nof its works while responding to the ChatGPT\u2019s users, in paragraph 15 of the\ninterim  application.  ANI  claims  these  examples  are  illustrative  and  not\nexhaustive.\n57.3. ChatGPT\u2019s  output  retains  the  original  and  creative  elements  of  the\nANI\u2019s  text  and  thereby  violates ANI\u2019s  proprietary  rights.  Some  of  the\nresponses demonstrate exact or nearly exact copies of ANI\u2019s data. This occurs\ndue to memorization of the raw data\/ training data.\n57.4. The  outputs  generated  by  ChatGPT  are  not  derivative  literary  works\neither,  as  it  requires  the  author  to  exercise  independent  skill,  labour,  and\n\n\n\nCS(COMM) 1028\/2024     Page 24 of 135\n\nresources  [Ref. Eastern  Book  Company v. D.B.  Modak\n## 18\n].  In  the  present\ncase,  the  user  generated  responses  show  that  there  is no  independent\napplication of mind. At best, the user generated responses can be considered\nadaptations of ANI\u2019s works as these responses are merely a re-arrangement\nof ANI\u2019s works.\n57.5. Open AI itself admits that verbatim reproduction of content from the\nraw data occurs in cases where the same content occurs repeatedly in the raw\ndata,  through  multiple  sources on the internet.  The  very  nature  of  ANI\u2019s\nbusiness model makes its copyrighted works available on various platforms\nincluding its subscribers\u2019 websites, which are leading news publications.\n57.6. The defence of Open AI that ANI\u2019s works are available in the public\ndomain and can be used for training Open AI\u2019s LLM without obtaining a\nlicense from ANI, is not valid and strikes at the root of copyright law. Public\navailability  does  not  nullify  copyright  or  grant  a  universal  license  for\nunrestricted use. ANI is not divested of its exclusive rights in its works only\nbecause it has licensed its content to its subscribers.\n57.7. Open  AI  has  recognized  similar  proprietary  rights  of  other  news\norganizations   and   entered   into   license   agreements   with   other   news\norganizations such as Financial Times, Associated Press and Cond\u00e9 Nast, and\nhas acknowledged the copyright of these entities.\n57.8. Open AI has also infringed ANI\u2019s exclusive right to translate the works,\nfor which ANI invests significant time and effort. In this regard, reliance is\nplaced on the English version of the interview given by Mr. Neeraj Chopra\u2019s\nmother,  which  was  originally in  Hindi  and  was  translated  into  English\n\n## 18\n## (2008) 1 SCC 1.\n\n\n\nCS(COMM) 1028\/2024     Page 25 of 135\n\nby ANI.\n57.9. In his rejoinder submissions, counsel for ANI placed reliance on the\nfollowing foreign judgments:\n(i) In The Associated Press v. Meltwater U.S. Holdings, Inc.\n## 19\n, the US\nDistrict Court held  that  summarizing  plaintiff\u2019s  works  using\ncomputer  programme  without  adding  any  commentary  is  not\ntransformative and hence, it does not constitute fair use under the\nUS Copyright Act.\n(ii) In GEMA v. Open AI\n## 20\n, the Munich Regional Court granted an\ninjunctive relief in favour of GEMA and against Open AI holding\nthat  there  is  memorization  of  training  data  and  consequential\nreproduction.\n(iii) In The Advance Local Media LLC et al. v. Cohere Inc.\n## 21\n, the US\nDistrict Court denied Cohere\u2019s partial motion to dismiss, holding\nthat substitutive summaries could constitute copyright infringement,\nand that the generation of hallucinated articles falsely attributed to\npublishers\/plaintiffs constituted a classic passing-off claim.\n\nIntervenors supporting ANI\ni. Submissions on behalf of the Intervenor \u2013 DNPA (I.A. 2199\/2025)\n&#8211; Mr. Rajshekhar Rao, senior counsel appearing on behalf of intervenors\nin I.A. 2199\/2025 made the following submissions:\n58.1. The  LLM  models  inherently  memorize  content.  This  can  be  implied\n\n## 19\n931 F. Supp. 2d 537 (S.D.N.Y. 2013).\n## 20\ncase no. 42 O 14139\/24, decided on 11\n\n## November 2025.\n## 21\nAdvance Local Media LLC v. Cohere Inc., 1:25-cv-01305, (S.D.N.Y.).\n\n\n\nCS(COMM) 1028\/2024     Page 26 of 135\n\nfrom the fact that LLMs can reproduce verbatim copies of unique forms of\ncontent on which they are trained.\n58.2. The  reproduction of copies of  training  data  implies  that  LLMs retain\nsuch data. The submission made on behalf of Open AI that LLMs only keep\nthe non-expressive components of copyrighted works is incorrect. It is also\nincorrect to suggest that LLM models only learn rules of grammar and syntax.\n58.3. The Copyright Act does not distinguish between \u2018expressive\u2019 and \u2018non-\nexpressive\u2019 uses for the purpose of copyright infringement.\n\nii. Submissions on behalf of the Intervenor &#8211; Indian Music Industry (I.A.\n## 4027\/2025)\n&#8211; Mr. C.M. Lall, senior counsel appearing on behalf of intervenors in I.A.\n4027\/2025 made the following submissions:\n59.1. Where the output amounts to substantial reproduction of original work,\nthe same would amount to copyright infringement. It is well-established that\nLLM models retain copies of their training data and this is apparent from the\nadmission made by Open AI with regard to \u2018regurgitation\u2019.\n59.2. Accordingly, regurgitation or extraction of training data from the LLM\nmodel, which results in a response\/ output to a user prompt being a substantial\nreproduction of the original work, will constitute infringement.\n\niii. Written Submissions filed by Mr. Ankit Sahni, Advocate on behalf of the\nIntervenor \u2013 Federation of Indian Publishers (FIP) (I.A. 470\/2025)\n&#8211; ChatGPT  regurgitates  works  on  which  it  has  been  trained  in  the\noutputs\/responses   generated   for   its   users   and   this   would   amount   to\ninfringement   under   the   Copyright   Act   as   the   outputs\/responses   are   a\nsubstantial reproduction of the copyrighted works.\n\n\n\nCS(COMM) 1028\/2024     Page 27 of 135\n\n\nSubmissions on behalf of the Defendant (OPEN AI)\n&#8211; Mr. Amit Sibal, senior counsel appearing on behalf of Open AI made\nthe following submissions:\n61.1. The cut-off date for training an LLM is several months prior to the\nlaunch of a model. In respect of the models available to ANI in the lead up to\nthe present suit, the cut-off dates for training were either April 2022 (for GPT\n4) or April 2024 (for GPT 4o). Each of the 14 (fourteen) works that ANI\nalleges  have  been  infringed  in  the  plaint  were  admittedly  authored  and\npublished subsequent to April 2024 (in either August or September 2024). It\nis therefore undisputed that the training data on which Open AI\u2019s LLMs were\ntrained  did  not  contain  the  14  (fourteen)  works  that  ANI  alleges  were\ninfringed.\n61.2. ANI has not identified any other works that it claims copyright over.\nTherefore,  ANI  cannot  seek  a  speculative  finding  of  infringement  over\nunidentified works.\n61.3. There is no exclusive right to \u2018use\u2019 any works under Section 14 of the\nCopyright Act. Therefore, ANI\u2019s allegation regarding \u2018use\u2019 of ANI\u2019s work by\nOpen AI cannot amount to infringement. The LLMs underlying ChatGPT do\nnot store or reproduce training data. Nor do they access data on which they\nwere trained once the pre-training phase is complete. The only \u2018use\u2019 for which\nthis data is deployed is during the training of Open AI\u2019s LLMs underlying\nChatGPT.\n61.4. ChatGPT is not designed to reproduce the extracts of content forming\na part of its training data, but may use information or facts that it has \u2018learnt\u2019\nfrom such content in the course of its training towards responses to the user\n\n\n\nCS(COMM) 1028\/2024     Page 28 of 135\n\nprompts. Therefore, ChatGPT cannot be said to \u2018duplicate\u2019 or \u2018reproduce\u2019 the\n\u2018manner  of  presentation\u2019 of  content  on  which  the  models  underlying\nChatGPT have been trained.\n61.5. It is the settled position of law that copyright holder\u2019s rights are limited\nto  the  protection  of  the  manner  in  which  an  idea  or  a  fact  is  expressed.\nHowever, there  is  no  such  protection  on  an  underlying  idea  or fact. This is\nespecially true in the context of news, where the underlying fact is incapable\nof copyright protection. In fact, there is a predominant public interest in the\nwide dissemination of such facts. Hence, ANI cannot claim monopoly of use\nover the said facts.\n61.6. Copyright in news and other factual literary content is limited only to\nthe  expression  thereof.  In  the  case  of \u2018news\u2019,  the  threshold  for  establishing\nsubstantial similarity in the expression is higher. [Ref: Feist Publications v.\n## Rural Telephone Service\n## 22\n## ].\n61.7. Grammatical  rules  and  syntax  by  themselves  cannot  be  protected  by\ncopyright because they are basic rules for sentence formation.\n61.8. Indian copyright law explicitly rejects the \u2018sweat of the brow\u2019 doctrine\nand has adopted the \u2018skill  and  judgment  test\u2019 to determine whether a work\nwould  be  entitled  to  protection  under  Indian copyright  law.  [Ref. Eastern\nBook Company v. D.B. Modak (supra)]\n61.9. ANI has sought to present limited extracts of a particular work rather\nthan the whole work to allege infringement. For the purpose of determining\ncopyright infringement, it is not permissible to split or compartmentalize or\n\n## 22\n## 499 U.S. 340 (1991).\n\n\n\nCS(COMM) 1028\/2024     Page 29 of 135\n\ndissect ANI\u2019s work into parts, then seek to compare such parts against the\nalleged infringed work. [Ref. Star India v. Leo Burnett\n## 23\n## ]\n61.10. ANI  has  not  even  been  able  to  demonstrate  substantial similarity,  let\nalone verbatim reproduction in respect of the limited instances that have been\ncited in the plaint. This is despite the fact that ANI made extreme efforts to\ntry  and  obtain  a  verbatim\/  substantially  similar  response  from  ChatGPT\nthrough a series of adversarial prompts.\n61.11. An  infringing  copy  must  be  a  substantial  or  material  reproduction.\nWhere  similarities  in  the  two  works  exist  alongside  broad  dissimilarities,\nwhich negate the intention to copy the original and the coincidences appearing\nin  the  two  works  are  clearly  incidental,  no  infringement  of  the  copyright\ncomes into existence. [Ref. R.G. Anand v. Deluxe Films\n## 24\n## ; Bikramjeet Singh\nBhullar v. Yash Raj Films\n## 25\n## ]\n61.12. Different ChatGPT users receive different responses based on the same\nprompts.\n61.13. Even in respect of ANI\u2019s allegations that the ChatGPT search feature\nreproduces results from ANI\u2019s website, it is submitted that a mere perusal of\nthe alleged response(s) relied upon by ANI makes it apparent that the same\nare  neither  identical  nor  substantially  similar  to  the  corresponding  articles\nrelied upon by ANI. Further, in such a case, ChatGPT provides a reference\/\nlink to the source but does not reproduce content therefrom.\n\n## 23\n2002 SCC OnLine Bom 942.\n## 24\n## (1978) 4 SCC 118.\n## 25\n(2023) SCC OnLine Del 8212.\n\n\n\nCS(COMM) 1028\/2024     Page 30 of 135\n\n61.14. \u2018Regurgitation\u2019 of data on which the LLM models are trained, is a rare\nglitch that is being actively addressed by Open AI. ANI has not been able to\nidentify any specific instance where \u2018regurgitation\u2019 of their works occurred.\n\nIntervenors supporting OPEN AI\ni. Submissions on behalf of the Intervenor &#8211; Indian Governance and Policy\nProject (IGAP) (I.A. 4616\/2025)\n&#8211; Mr. Akhil Sibal, senior counsel appearing on behalf of the intervenor\nin I.A. 4616\/2025 made the following submissions:\n62.1. In Andrea Bartz v. Anthropic\n## 26\n, the US District Court emphasized that\nwhat  is  relevant  is  not  the  amount  of  data  used  for  making  a  copy,  but  the\namount  and  substantiality  of  what  is  made  accessible  to  the  public  by  the\nsecondary use. Since LLMs do not regurgitate an author\u2019s work or produce\nsubstantially similar works, it would not amount to copyright infringement.\n62.2. An LLM model does not store the training data and memorization of\ntraining  data  is  a  rare  occurrence.  Hence,  the  display  at  the  output  creation\nstage is a result of manipulation of the system by the user alone.\n\nii. Submissions on behalf of the Intervenor &#8211; Broadband India Forum (BIF)\n## (I.A. 10766\/2025)\n&#8211; Mr. Kapil Sibal and Mr. Arvind P. Datar, senior counsel appearing on\nbehalf of the intervenor in I.A. 10766\/2025 made the following submissions:\n63.1. Newspaper reports, including articles which are in the public domain\nare not reproduced in the output in any material form by LLMs. If there is a\nprompt  on  a  particular  event,  the  LLM  is  able  to  examine  the  voluminous\n\n## 26\nBartz v. Anthropic PBC, 3:24-cv-05417, (N.D. Cal.).\n\n\n\nCS(COMM) 1028\/2024     Page 31 of 135\n\ninformation  available  in  the  public  domain  and  prepare  a  unique  response.\nTherefore,  there  is  no  reproduction  in  any  material  form.  The  LLMs  often\nprovide links to the sources that have been accessed.\n63.2. Several  newspapers,  magazines,  etc.  have  created  \u2018paywalls\u2019  or\n\u2018crawler  blockers\u2019 to prevent access to their contents. The data from these\nwebsites  is  completely  blocked  and  cannot  be  accessed.  Preparing  the\nresponses based on publicly available websites will not amount to violation\nof Section 14(a)(i) of the Copyright Act, so long as there is no reproduction\nin  material  form.  LLMs  do  not  reproduce  the  literary  work  in  any  material\nform.\n63.3. LLMs  are  not  designed  to  materially  reproduce  any  work  or  a\nsubstantial part of it in their response to prompts. They merely understand the\nideas, concepts and the facts in those works, which can be used to generate\noutput. It does not amount to adaptation of copyrighted works.\n\niii. Submissions on behalf of the Intervenor \u2013 Flux Labs Pvt. Ltd. (I.A.\n## 4615\/2025)\n&#8211; Ms.  Haripriya  Padmanabhan, senior  counsel appearing  on  behalf  of\nintervenor in I.A. 4615\/2025 made the following submissions:\n64.1. The  examples  given  by  ANI  in  the  plaint  do  not  show  substantial\nreproduction of copyrighted works of ANI.\n64.2. Unlike  any  other  customer  of  an  AI  platform,  ANI  gave  a  \u2018direct\nprompt\u2019  to  ChatGPT  in  the  said  examples.  Yet,  the  responses  given  by\nChatGPT only summarised the interviews in its own words and referred to the\nauthor in third person.\n64.3. The  AI  platforms  also  provide  a  reference  and  do  not  claim  that  the\n\n\n\nCS(COMM) 1028\/2024     Page 32 of 135\n\ninformation is theirs. Hence, the output\/responses generated by AI platforms\ncannot be stated to be a violation of copyright under Section 14(a)(i) of the\n## Copyright Act.\n\nSubmissions of Mr. Adarsh Ramanujan, Amicus Curiae\n&#8211; Mr. Ramanujan made the following submissions:\n65.1. The issue whether Open AI\u2019s LLMs underlying ChatGPT retain the full\n\u2018raw data\u2019 and reproduce duplicate or near duplicate extracts of the \u2018raw data\u2019\nas alleged by ANI can be determined only after a full trial. However, ANI\u2019s\nsubmission  in  this  regard  does  not  appear  to  be prima  facie tenable  as  the\nfunction  of  an  LLM  is  based  on  patterns  and  therefore,  it  would  not  make\nsense for an LLM to store raw data on a permanent basis to generate responses\nfor a query.\n65.2. In  some  instances,  training  of  an  LLM  leads  to  memorization.  This\ntendency of verbatim memorization occurs when an LLM is trained on basic\nfacts  or  unique  pieces  of  information,  since  the  training  data  may  not  be\nsufficiently large for the model to build a generalization. Hence, it is possible\nthat  training  of  ChatGPT  leads  the  underlying  LLMs  to  memorize  parts  of\nANI\u2019s works.\n65.3. In   the   event   memorization   occurs,   even   if   inadvertently,   the\ncommunication of the same to the public at the output stage would amount to\ninfringement.\n65.4. The decision  of the  Supreme Court in R.G. Anand (supra),  is  the\nleading  judgment  on  the  aspect  of  copyright  infringement.  In  terms  of\nparagraph 46 of R.G. Anand (supra), the Supreme Court has clearly stated\n\n\n\nCS(COMM) 1028\/2024     Page 33 of 135\n\nthat \u201cform, manner and arrangement and expression of the idea\u201d would only\nbe entitled to copyright protection.\n65.5. Retrieval Augmented Generation (RAG)\n## 27\nis an aspect which has not\nbeen pleaded in the plaint and has been raised only in the course of arguments.\nTherefore, whether RAG involves infringement of ANI\u2019s works should also\nbe a subject matter of a trial.\n\nSubmissions of Professor Arul George Scaria, Amicus Curiae\n&#8211; Prof. Scaria made the following submissions:\n66.1. A  reading  of  the  judgment  by  the  Supreme Court in Eastern  Book\nCompany v. D.B.  Modak (supra),  highlights  that  India  has  explicitly\ndiscarded  the \u2018sweat  of  the  brow\u2019 approach  and  adopted  the \u2018skill  and\njudgment\u2019 standard. Therefore, the current legal standard of originality under\nIndian copyright  law  is  of  a  standard  higher  than  the \u2018sweat  of  the  brow\u2019\nstandard.\n66.2. Taking  into  account  the  current  standard  of  originality,  substantial\nchunk  of  news  related  materials  might  be  outside  copyright  protection  in\nIndia. Therefore, issues such as existence of a copyright in favour of ANI and\nthe need for use of news material by LLMs to provide accurate responses can\nonly be determined in a trial.\n66.3. The  non-expressive  use  of  copyrighted  works  by  LLMs  to  provide\nmore  optimal  responses  to  any  user  prompts  may  not  constitute  copyright\ninfringement.\n66.4. Since news is based on facts and when there are limited ways in which\n\n## 27\nAs discussed in the Section: Working of LLMs.\n\n\n\nCS(COMM) 1028\/2024     Page 34 of 135\n\na  fact  can  be  expressed, in  terms  of \u2018merger  doctrine\u2019  as  discussed  in\nSyndicate of the Press of the University of Cambridge v. B.D. Bhandari\n## 28\n## ,\nnews should not get copyright protection or should get only a thin layer of\nprotection.\n## DISCUSSION AND ANALYSIS\n&#8211; Under the Copyright Act, copyright subsists only for the classes of\nworks enlisted under Section 13 of the Copyright Act. ANI claims that news\narticles and interviews published on its digital platforms are \u201coriginal literary\nworks\u201d under Section 13(1)(a)\n## 29\nof the Copyright Act. This position has not\nbeen disputed by Open AI. However, Open AI submits that ANI is not the\nowner  of  works  published  on  its  website  in  terms  of  Section  17\n## 30\nof  the\n## Copyright Act.\n&#8211; To establish ownership of the works published on its website, ANI has\nplaced  on  record  a  sample  \u2018Professional Services Agreement\u2019  in  terms  of\nwhich copyright would vest with ANI in respect of original works, such as\nimages,   video   clippings,   articles,   reports,   interviews,   etc.\ncreated\/authored\/obtained by personnel on behalf of ANI.\n\n## 28\n2011 SCC OnLine Del 3215.\n## 29\n&#8211; Works in which copyright subsists.\u2014(1) Subject to the provisions of this section and the other\nprovisions of this Act, copyright shall subsist throughout India in the following classes of works, that is to\nsay,\u2014\n(a) original literary, dramatic, musical and artistic works;\n\n## 30\n&#8211; First owner of copyright.\u2014Subject to the provisions of this Act, the author of a work shall be the first\nowner of the copyright therein:\nProvided that\u2014\n(a) in the case of a literary, dramatic or artistic work made by the author in the course of his employment by\nthe proprietor of a newspaper, magazine or similar periodical under a contract of service or apprenticeship,\nfor the purpose of publication in a newspaper, magazine or similar periodical, the said proprietor shall, in the\nabsence of any agreement to the contrary, be the first owner of the copyright in the work, in so far as the\ncopyright relates to the publication of the work in any newspaper, magazine or similar periodical, or to the\nreproduction of the work for the purpose of its being so published, but in all other respects the author shall\nbe the first owner of the copyright in the work;\n\n\n\nCS(COMM) 1028\/2024     Page 35 of 135\n\n&#8211; On a prima facie view, this Court is satisfied that there is an explicit\nagreement between ANI and its professionals which recognises ANI as the\nowner of the copyright subsisting in \u201coriginal literary works\u201d published on\nits website in terms of Section 17 of the Copyright Act.\n&#8211; It is not disputed by Open AI that \u201coriginal literary works\u201d that are\navailable  free  of  cost  and  publicly  accessible  are  nevertheless  entitled  to\ncopyright  protection.  Therefore,  for  adjudicating  present  application  this\nCourt is of the prima facie view that ANI, as an owner, would have copyright\nover  the  \u201coriginal  literary  works\u201d,  even  if  they  are  freely  and  publicly\navailable on its website. Any challenge to the ownership of individual works\nwhich are subject matter of the present suit can be adjudicated at the stage of\ntrial.\n&#8211; As a copyright owner of the original literary works, ANI enjoys certain\nexclusive rights under Section 14 of the Copyright Act, which include the\nfollowing:\n(i) the  exclusive  right  to  reproduction  of  the  work  in  terms  of\nSection (14)(a)(i) of the Copyright Act and,\n(ii) the right to \u201ccommunicate it to the public\u201d in terms of Section\n14(a)(iii) of the Copyright Act.\nFor the ease of reference, the relevant extracts from Section 14 are set out\nbelow:\n\n\u201c14. Meaning of copyright.\u2014For the purposes of this Act, \u201ccopyright\u201d\nmeans the exclusive right subject to the provisions of this Act, to do or\nauthorise the doing of any of the following acts in respect of a work or\nany substantial part thereof, namely:\u2014 (a) in the case of a literary,\ndramatic or musical work, not being a computer programme,\u2014\n\n\n\nCS(COMM) 1028\/2024     Page 36 of 135\n\n(i) to reproduce the work in any material form including the storing of\nit in any medium by electronic means;\n## ***   ***   ***\n(iii) to perform the work in public, or communicate it to the public;\u201d\n## 72. Section 2 (ff)\n## 31\nof the Copyright Act defines \u201ccommunication to the\npublic\u201d.\n&#8211; Thus, the act of reproduction and communicating the work of ANI to\nthe public by a third party without license would amount to infringement as\nper Section 51\n## 32\nof the Copyright Act\n&#8211; It is the case of ANI that Open AI\u2019s LLM model memorizes the content\nof ANI\u2019s articles and specific responses are generated by ChatGPT through\nmemorisation and the original literary works of ANI are communicated to the\npublic. This amounts to infringement under Section 51 of the Copyright Act.\n\n## 31\n(ff) \u201ccommunication to the public\u201d means making any work or performance available for being seen or\nheard or  otherwise  enjoyed  by  the  public  directly  or  by  any  means  of  display or diffusion  other  than  by\nissuing physical copies of it, whether simultaneously or at places and times chosen individually, regardless\nof whether any member of the public actually sees, hears or otherwise enjoys the work or performance so\nmade available.\n## 32\n&#8211; When copyright infringed.\u2014Copyright in a work shall be deemed to be infringed\u2014\n(a) when any person, without a licence granted by the owner of the Copyright or the Registrar of Copyrights\nunder this Act or in contravention of the conditions of a licence so granted or of any condition imposed by a\ncompetent authority under this Act\u2014\n(i) does anything, the exclusive right to do which is by this Act conferred upon the owner of the\ncopyright, or\n(ii) permits for profit, any place to be used for the communication of the work to the public where such\ncommunication constitutes an infringement of the copyright in the work, unless he was not aware and\nhad no reasonable ground for believing that such communication to the public would be an infringement\nof copyright; or\n(b) when any person\u2014\n(i) makes for sale or hire, or sells or lets for hire, or by way of trade displays or offers for sale or hire,\nor\n(ii) distributes either for the purpose of trade or to such an extent as to affect prejudicially the owner of\nthe copyright, or\n(iii) by way of trade exhibits in public, or\n(iv) imports into India,\nany infringing copies of the work:\nProvided that nothing in sub-clause (iv) shall apply to the import of one copy of any work for the private and\ndomestic use of the importer.\nExplanation.\u2014For the purposes of this section, the reproduction of a literary, dramatic, musical or artistic\nwork in the form of a cinematograph film shall be deemed to be an \u201cinfringing copy\u201d.\n\n\n\nCS(COMM) 1028\/2024     Page 37 of 135\n\n&#8211; Therefore, this Court would have to examine whether the responses\ngenerated by ChatGPT would result in reproduction and communication of\nthe original literary works of ANI to the public without authorisation from\nANI and hence, would amount to infringement. More particularly, the Court\nwould have to examine the following two issues:\ni. whether Open AI memorizes and regurgitates ANI\u2019s copyrighted\nliterary works in the form of responses and,\nii. whether the ChatGPT\u2019s responses are substantial reproduction of\nANI\u2019s copyrighted literary works.\n\ni. Whether  Open AI  memorizes  and  regurgitates  ANI\u2019s  copyrighted\nliterary works in the form of responses?\n\n&#8211; It is contended on behalf of ANI and the intervenors supporting ANI\nthat  LLM  models  underlying  ChatGPT  inherently  memorize  content  and\nretain copies of their training data which results in regurgitation. According\nto ANI and intervenors, LLMs are capable of providing verbatim or near-\nidentical  reproductions  of  the  training  data  in  response  to  user  prompts,\nthereby demonstrating that expressive elements of the copyrighted works are\nalso retained and reproduced.\n&#8211; Per contra, it has been highlighted on behalf of Open AI as well as\nintervenors  supporting  Open  AI  that  an  LLM  model  is  not  designed  to\nreproduce extracts of the contents that form part of their training data. It was\nalso argued that the LLMs underlying ChatGPT are not trained to duplicate or\nregurgitate the exact manner of presentation of content on which these models\nhave been trained. Based on a particular prompt, the LLM is able to prepare\na unique response based on the voluminous data on which it has been trained.\n\n\n\nCS(COMM) 1028\/2024     Page 38 of 135\n\nFurther, generative AI models also provide links to the sources that have been\naccessed  by  them.  In  this  regard,  Open AI  has highlighted that the  same\nprompts  by  different  users  would  result  in  different  responses  from  the\nChatGPT (document 10 in the defendant\u2019s documents filed along with the\nwritten statement).\n&#8211; Professor Scaria, one of the Amici Curiae, in his written submissions\nhas argued that in vast majority of cases there is minimal memorisation in the\nworking of LLMs and mostly memorisation occurs when specific attempts\nare  made  to  generate  potential  copyright  infringing  outputs. The  relevant\nextracts from the written submissions of Prof. Scaria are set out below:\n\u201c58. A couple of general observations may be added here. First, as the\ntechnical literature suggests, though LLMs may be using copyrighted\ntraining data to generate outputs, in most instances, they may not be\ngenerating  potentially  copyright  infringing  responses.  Scholars  like\nProf. Matthew Sag points out that this is due to the different steps taken\nduring  model  training  like  the  process  of  decomposition  and\nabstraction,  blending  of  latent  concepts,  and  noise  injection  [Sag,\n2023]. However,  in  some  instances,  particularly  due  to  the\nmemorisation problem discussed earlier as part of Issue II, there may\nbe outputs which are identical or similar to the input (copyrighted)\ntraining data. This is seen mostly during the so-called \u201cextraction\nattacks\u201d, wherein specific attempts are made to generate potentially\ncopyright infringing outputs. Some of the empirical studies suggest\nthat  memorisation  was  seen  in  around  0.03%  of  sample  images\nduring extraction attacks and they were successful mostly in instances\nwherein the images were duplicated more than 100 times [Sag, 2023].\nThis also suggests that in the vast majority of use scenarios, the use\nof copyrighted materials for generating responses are non-infringing,\nbecause  only  non-expressive  uses are  made  of  the  copyrighted\nmaterials.\u201d\n[Emphasis supplied]\n\n&#8211; Mr. Adarsh  Ramanujan  in  his  written  submissions  has  argued  that\n\n\n\nCS(COMM) 1028\/2024     Page 39 of 135\n\nANI\u2019s  contention  that  Open AI  retains  the  full  raw  data  is prima  facie\nuntenable as an LLM is merely a set of mathematical functions to connect\ninput  to  output. An  LLM\u2019s  function  is  to  predict  based  on  patterns.  The\nchances of memorization of the raw data may increase when an LLM Model\nis trained on basic facts or unique pieces of information. This tendency occurs\nbecause  there  may  be  absence  of  general  patterns  leading  to  direct\nreproduction at the time of output generation.\n&#8211; The US District Court for Northern District of California in Bartz v.\nAnthropic (supra) examined  the  issue  of  memorization  in  LLMs  and\nregurgitation  of  responses  from  its  training  data  and  made  the  following\nobservations:\n\u201cEither way Meta added the books it downloaded to the datasets it used\nto train the Llama models. It is also post-trained its models to prevent\nthem from \u201cmemorizing and outputting certain text from their training\ndata,  including  copyrighted  material.  These  training  efforts,  which\nMeta  calls  \u201cmitigations,\u201d  appear  to  have  been  successful. Meta\u2019s\nexpert witness tested them using a method designed to get LLMs to\nregurgitate  material  from  its  training  data  (which  Meta  calls\n\u201cadversarial prompting\u201d). Even using that method, the expert could\nget no model to generate more than 50 words and punctuation marks\n(that is, \u201ctokens\u201d) from the plaintiffs\u2019 books. And the plaintiffs\u2019 expert\ncould only get the Llama model best at regurgitation to generate 50\nwords and punctuation marks from the plaintiffs\u2019 books in 60% of\ntests. She also testified that Llama was not able to reproduce \u201cany\nsignificant percentage\u201d of them. Meta MSJ Ex.24 at 237:16-19:see\nalso Pls Ex. 79, 70-72, 79, 82-83, 92:Meta MSJ Ex.23 at 179:22-25,\n180:17-181:16. In short, Llama cannot currently be used to read or\nother wise meaningfully access the plaintiffs\u2019 books.\u201d\n[Emphasis supplied]\n\n&#8211; From the aforesaid observations, it appears that even the plaintiffs\u2019\nexpert  in  the  field  of  AI  could  only  make  the  LLM  Model  regurgitate\n\n\n\nCS(COMM) 1028\/2024     Page 40 of 135\n\ninsignificant amounts from the copyrighted works of the plaintiffs.\n&#8211; ANI\u2019s case is based on \u2018memorisation\u2019, whereby the LLM memorises\nthe training data verbatim or substantially. In the plaint, ANI has given nine\nillustrations  to  demonstrate  substantial  reproduction.  It  is  an  undisputed\nposition that the training of the LLMs underlying ChatGPT ended in April\n2022 (for GPT 4) or April 2024 (for GPT 4o), which was much prior to the\npublication of the illustrative articles used in the plaint, to allege reproduction\nof ANI\u2019s copyrighted works. The table below illustrates this aspect:\n## TITLE(S) OF THE ARTICLE(S) DATE OF PUBLICATION\n\u201cMy  father  was  forced&#8230;\u201d:  Omar\nAbdullah opens up on IC 814 hijack\nincident,  reveals  difficult  decisions\nfaced by his father\n## 6\nth\n## September 2024\nT20 World Cup: \u201cPakistan never  a\nchallenge in Women\u2019s cricket\u201d: Ex-\nplayer Reema Malhotra backs India\n## 30\nth\n## September 2024\n\u201cLord  Balaji  will  not  forgive  him\neven if he climbs up the seven hills\non  his  knees\u201d:  TDP  spokesperson\nslams Jagan Mohan Reddy\n## 26\nth\n## September 2024\n\u201cConnecting  this  to  UP  not  right,\u201d\nHimachal  AICC  in-charge  Rajeev\nShukla  defends  state  govt  on  ID\ncards for vendors\n## 26\nth\n## September 2024\n\u201cIt\u2019s  gone  underground&#8230;\u201d:  JKPC\nchairman  Sajad  Gani  Lone  on\n\u2018Azadi\u2019 sentiments in Kashmir\n## 10\nth\n## September 2024\n\n\n\nCS(COMM) 1028\/2024     Page 41 of 135\n\n\u201cJinnah  like  mentality,\u201d  Minister\nHardeep Puri accuses Rahul Gandhi\nof  crossing  limit  for  \u201cpolitical\ninterest\u201d\n## 17\nth\n## September 2024\nNeeraj  Chopra\u2019s  mother  celebrates\nOlympic  Silver;  praises  Pakistan\u2019s\nArshad Nadeem for bagging Gold\n## 9\nth\n## August 2024\n\u201cAshwin best in India, Lyon better\noverall\u201d: Monty Panesar sets-up best\nTest spinner debate\n## 23\nrd\n## September 2024\n\n&#8211; Therefore, it is apparent that the illustrations of ChatGPT responses\nrelied upon in the plaint could not have been produced on account of training\nof Open AI\u2019s LLMs on the ANI\u2019s literary works, as the literary works in\nquestion were published only after the completion of the training process.\n&#8211; As discussed earlier under the Section \u2018Working of LLMs\u2019, it could be\ninferred that when an LLM refers to information on which it was not trained,\nit is using RAG technique. Using the RAG technique, LLMs retrieve relevant\ninformation  from  external  sources,  in  this  case  ANI\u2019s  website,  before\ngenerating  a  response.  Relevant  data  retrieved  from  the  website,  which  is\ncontextually aligned with the user\u2019s prompt, is then generated as  responses\ngrounded in the website\u2019s data rather than relying solely on its pre-trained\nknowledge. Therefore, responses that are not based on the trained data cannot\nbe used against Open AI in respect of claims based on memorisation. Such\nallegations  must  be  tested  based  on  the  identical  or  substantially  similar\nreproduction of the trained data. Therefore, this Court is of the prima facie\nview  that  the  instances  given  in  the  plaint  alleging  infringement  are  not  a\n\n\n\nCS(COMM) 1028\/2024     Page 42 of 135\n\nresult  of  memorisation, rather they are in the nature of live links, perhaps\nreflecting  RAG technique.  Whether  outputs  produced  using  RAG  would\namount to copyright infringement is an aspect which has not been pleaded in\nthe plaint, though it was referred during the course of submissions.\n\nii. Whether ChatGPT\u2019s responses are substantial reproduction of ANI\u2019s\ncopyrighted literary works?\n&#8211; To begin with, it may be relevant to discuss legal principles regarding\ncopyright infringement elucidated in some of the instructive judgments of the\nSupreme Court that have been relied upon by both sides.\n&#8211; In R.G. Anand (supra), the plaintiff sued the defendants for damages\nclaiming infringement of the copyright of the plaintiff in a dramatic play,\nwhich was made into a motion picture by the defendants. The Supreme Court\nheld  that  an  idea,  principle,  theme  or  facts  cannot  be  subject  matter  of\ncopyright protection. Where two writers write on the same subject, similarities\nare bound to occur because the central idea of both are the same. Therefore,\nthe Court has to determine whether the defendant has only adopted the idea\nof the copyrighted work or has also adopted the form, manner, arrangement\nand expression of an idea with minor changes. If the defendant\u2019s work appears\nto be a transparent rephrasing or a copy of the substantial and material part of\nthe original, it would amount to copyright infringement. The Supreme Court\nin paragraph 46 of the judgment has laid down the following principles to\ndetermine infringement of copyright, which are set out below:\n\u201c46. Thus, on a careful consideration and elucidation of the various\nauthorities  and  the  case  law  on  the  subject  discussed  above,  the\nfollowing propositions emerge:\n\n\n\nCS(COMM) 1028\/2024     Page 43 of 135\n\n&#8211; There can be no copyright in an idea, subject-matter, themes, plots\nor historical or legendary facts and violation of the copyright in such\ncases  is  confined  to  the  form,  manner  and  arrangement  and\nexpression of the idea by the author of the copyrighted work.\n&#8211; Where the same idea is being developed in a different manner, it is\nmanifest that the source being common, similarities are bound to\noccur. In such a case the courts should determine whether or not the\nsimilarities are on fundamental or substantial aspects of the mode of\nexpression adopted in the copyrighted work. If the defendant\u2019s work\nis nothing but a literal imitation of the copyrighted work with some\nvariations  here  and  there  it  would  amount  to  violation  of  the\ncopyright. In other words, in order to be actionable the copy must be\na substantial and material one which at once leads to the conclusion\nthat the defendant is guilty of an act of piracy.\n&#8211; One of the surest and the safest test to determine whether or not\nthere has been a violation of copyright is to see if the reader, spectator\nor the viewer after having read or seen both the works is clearly of the\nopinion and gets an unmistakable impression that the subsequent\nwork appears to be a copy of the original.\n&#8211; Where the theme is the same but is presented and treated differently\nso that the subsequent work becomes a completely new work, no\nquestion of violation of copyright arises.\n&#8211; Where however apart from the similarities appearing in the two\nworks there are also material and broad dissimilarities which negative\nthe intention to copy the original and the coincidences appearing in the\ntwo works are clearly incidental no infringement of the copyright comes\ninto existence.\n&#8211; As a violation of copyright amounts to an act of piracy it must be\nproved by clear and cogent evidence after applying the various tests\nlaid down by the case-law discussed above.\n&#8211; Where however the question is of the violation of the copyright of\nstage play by a film producer or a director the task of the plaintiff\nbecomes more difficult to prove piracy. It is manifest that unlike a stage\nplay a film has a much broader prospective, wider field and a bigger\nbackground  where  the  defendants  can  by  introducing  a  variety  of\nincidents give a colour and complexion different from the manner in\nwhich the copyrighted work has expressed the idea. Even so, if the\nviewer after seeing the film gets a totality of impression that the film\nis by and large a copy of the original play, violation of the copyright\nmay be said to be proved.\u201d\n[Emphasis supplied]\n\n\n\nCS(COMM) 1028\/2024     Page 44 of 135\n\n\n&#8211; In Eastern  Book  Company v. D.B.  Modak (supra),  the\nappellants\/plaintiffs filed  a  suit  along  with  an  application  for  temporary\ninjunction seeking to restrain the respondents\/defendants from reproducing\nthe plaintiffs\u2019 version of copy-edited judgments and orders of the Supreme\nCourt on CD-ROMs. The defendants were selling software packages on CD-\nROMs with the text of the copy-edited judgments of the Supreme Court along\nwith headnotes prepared by the plaintiffs for its own law report, \u2018SCC\u2019. The\nSupreme Court, relying upon the judgment of the Supreme Court of Canada\nin CCH Canadian v. Law Society of Upper Canada\n## 33\nheld that \u2018sweat of the\nbrow\u2019 approach to originality is not the correct standard as it fails to protect\nthe public interest in production and dissemination of works. On the other\nhand, the creativity standard of originality as applied by the US Courts, is a\nstandard of originality with a higher threshold. The underlying impetus for\nthis modification was the need to balance the rights of the copyright owner\nwith the public interest.\n&#8211; In paragraph 15 of the Eastern Book Company v. D.B. Modak (supra),\nit was specifically observed that for a literary work to be original and to claim\ncopyright, it should possess some amount of creativity in the work, more than\nmere facts. For ease of reference, the aforesaid paragraph is set out below:\n\u201c15.  Copyright  is  a  right  to  stop  others  from  exploiting  the  work\nwithout the consent or assent of the owner of the copyright. A copyright\nlaw presents a balance between the interests and rights of the author\nand that of the public in protecting the public domain, or to claim the\ncopyright and protect it under the copyright statute. One of the key\nrequirements is that of originality which contributes, and has a direct\nnexus, in maintaining the interests of the author as well as that of\n\n## 33\n## (2004) 1 R.C.S. 339.\n\n\n\nCS(COMM) 1028\/2024     Page 45 of 135\n\npublic in protecting the matters in public domain. It is a well-accepted\nprinciple of copyright law that there is no copyright in the facts per\nse, as the facts are not created nor have they originated with the\nauthor of any work which embodies these facts. The issue of copyright\nis closely connected to that of commercial viability, and commercial\nconsequences and implications.\u201d\n[Emphasis supplied]\n\n&#8211; Accordingly, the Supreme Court laid down the \u2018skill and judgment test\u2019\nas opposed to \u2018sweat of the brow\u2019 test that was applied earlier in India. In terms\nof the \u2018skill and judgment test\u2019, the author of a derivative work has to satisfy\nthat the work has been produced from exercise of skill and judgment and, the\nskill and judgment should not be of a trivial nature. It was observed that the\n\u2018sweat of the brow\u2019 doctrine creates a monopoly in materials available in\npublic domain without encouraging the creation of writings by the authors. In\nthis regard, the relevant paragraph of the Eastern Book Company v. D.B.\nModak (supra) is set out below:\n\n\u201c57. The Copyright Act is not concerned with the original idea but\nwith the expression of thought. Copyright has nothing to do with\noriginality  or  literary  merit.  Copyrighted  material  is  that  what  is\ncreated by the author by his own skill, labour and investment of\ncapital,  maybe  it  is  a  derivative  work  which  gives  a  flavour  of\ncreativity. The  copyright  work  which  comes  into  being  should  be\noriginal  in  the  sense  that  by  virtue  of  selection,  coordination  or\narrangement  of  pre-existing  data  contained  in  the  work,  a  work\nsomewhat different in character is produced by the author. On the face\nof the provisions of the Copyright Act, 1957, we think that the principle\nlaid down by the Canadian Court would be applicable in copyright of\nthe judgments of the Apex Court. We make it clear that the decision of\nours would be confined to the judgments of the courts which are in the\npublic domain as by virtue of Section 52 of the Act there is no copyright\nin  the  original  text  of  the  judgments. To  claim  copyright  in  a\ncompilation, the author must produce the material with exercise of\n\n\n\nCS(COMM) 1028\/2024     Page 46 of 135\n\nhis skill and judgment which may not be creativity in the sense that it\nis novel or non-obvious, but at the same time it is not a product of\nmerely  labour  and  capital.  The  derivative  work  produced  by  the\nauthor must have some distinguishable features and flavour to raw\ntext of the judgments delivered by the court. The trivial variation or\ninputs put in the judgment would not satisfy the test of copyright of\nan author.\u201d\n[Emphasis supplied]\n\n&#8211; The legal position that emerges from Eastern Book Company v. D.B.\nModak (supra) is that the work which is produced by the author by using his\nskill  and  judgement  with  a  flavour  of  creativity  is  copyrightable.  The\nprotection would be only in respect of creative and original works. There\ncannot be any copyright protection in respect of facts even if someone has put\nin effort to obtain those facts.\n&#8211; In Feist Publications (supra), the US Supreme Court laid down the test\nof \u2018modicum of creativity\u2019. In this case, a telephone utility company brought\na copyright infringement action against the publisher of a telephone directory.\nThe US Supreme Court held that no one can claim originality as to the facts.\nA person who reports a particular fact has not created the said fact but has\nonly discovered its existence. The relevant paragraphs of Feist Publications\n(supra) are set out below:\n\n\u201c[15] It is this bedrock principle of copyright that mandates the law\u2019s\nseemingly disparate treatment of facts and factual compilations. \u201cNo\none may claim originality as to facts.\u201d Id., \u00a7 2.11[A], p. 2-157. This is\nbecause facts do not owe their origin to an act of authorship. The\ndistinction is one between creation and discovery : the first person to\nfind and report a particular fact has not created the fact; he or she\nhas merely discovered its existence. To borrow from Burrow-Giles, one\nwho discovers a fact is not its \u201cmaker\u201d or \u201coriginator.\u201d 111 U.S., at 58.\n\u201cThe discoverer merely finds and records.\u201d Nimmer \u00a7 2.03[E]. Census-\n\n\n\nCS(COMM) 1028\/2024     Page 47 of 135\n\ntakers, for example, do not \u201ccreate\u201d the population figures that emerge\nfrom their efforts; in a sense, they copy these figures from the world\naround them. Denicola, Copyright in Collections of Facts : A Theory\nfor the Protection of Nonfiction Literary Works, 81 Colum. L. Rev. 516,\n525  (1981)  (hereinafter  Denicola). Census  data  therefore  do  not\ntrigger  copyright  because  these  data  are  not  \u201coriginal\u201d  in  the\nconstitutional sense. Nimmer [p*348] \u00a7 2.03 [E]. The same is true of\nall facts &#8212; scientific, historical, biographical, and news of the day.\n\u201cThey may not be copyrighted and are part of the public domain\navailable to every person.\u201d Miller, supra, at 1369.\n## ***   ***   ***\n[18] This inevitably means that the copyright in a factual compilation\nis  thin.  Notwithstanding  a  valid  copyright,  a  subsequent  compiler\nremains free to use the facts contained in an another&#8217;s publication to\naid in preparing a competing work, so long as the competing work does\nnot feature the same selection and arrangement. As one commentator\nexplains  it:\u201cNo  matter  how  much  original  authorship  the  work\ndisplays, the facts and ideas it exposes are free for the taking&#8230;. The\nvery same facts and ideas may be divorced from the context imposed\nby the author, and restated or reshuffled by second comers, even if the\nauthor was the first to discover the facts or to propose the ideas.\u201d\n\u201cGinsburg 1868.\u201d\n[Emphasis supplied]\n\n&#8211; ANI  has  relied upon  judgment of the  House  of  Lords in Walter v.\n## Lane\n## 34\ndelivered in 1900, to show that it has copyright in respect of the news\ncirculated by its platform. In Walter v. Lane (supra), the issue before the\nHouse of Lords was whether a person who makes notes of speech delivered\nin public,  transcribes  the same  and publishes in the  newspaper  would  be\nentitled to copyright protection. The House of Lords held that each of the\nreporters reporting the speech would have a copyright in his own published\nreport. The House of Lords came to this conclusion based on the doctrine of\n\n## 34\n## [1900] AC 539.\n\n\n\nCS(COMM) 1028\/2024     Page 48 of 135\n\n\u2018sweat of the brow\u2019 and it was premised on the principle that an author should\nbe able to reap the benefits of its labour.\n&#8211; On  behalf  of  Open  AI,  the  aforesaid  judgment  is  sought  to  be\ndistinguished as it was in the context of copyright laws that existed in England\nin 1842. The then copyright laws in England did not require originality.\n&#8211; Walter v. Lane (supra)  was  considered  by  the  Supreme Court in\nEastern Book Company v. D.B. Modak (supra), as the appellant therein relied\nupon the same, and it was held that the doctrine of \u2018sweat of the brow\u2019 does\nnot hold good anymore as it gives more weight to the copyright owner\u2019s right\nas opposed to the public interest.\n&#8211; In B.D. Bhandari (supra), the Division Bench delved on the \u2018merger\ndoctrine\u2019 and noted that when there is limited manner in which a fact can be\nexpressed,  the  facts  would  merge  with  the  expression. The  facts  and  the\nexpression would become so intrinsically connected that it would be difficult\nto  distinguish  one  from  another,  making  the  expression  itself  non-\ncopyrightable.  For  ease  of  reference,  the  relevant  paragraphs  of B.D.\nBhandari (supra) are set out below:\n\u201c106. While  answering  the  first  question,  the  learned  Single  Judge\nexpanded  the  \u201cthe  Idea-Expression  Dichotomy  and  the  Doctrine  of\nMerger\u201d and noted that there is no copyright protection on ideas and\nfacts and it is only the creative expression of such ideas and facts that is\nrewarded by law. In the process, expression \u201coriginality\u201d in artistic,\nliterary, dramatic and musical works were also explained holding that\nwith the passage of time, doctrine of \u201csweat of the brow\u201d had given way\nto the \u201cmodicum of creativity\u201d and therefore, in order to have a copyright\nin a work, it was necessary to show that it involves a certain degree of\ncreativity.\n&#8211; The learned Single Judge also  deemed it proper to  refer to  the\n\u201cdoctrine of merger\u201d, peculiar to copyright law in the following manner:\n\n\n\nCS(COMM) 1028\/2024     Page 49 of 135\n\n\u201c18. An aspect peculiar to copyright law, is the doctrine of\nmerger, is involved in this case. This doctrine posits that where\nthe idea and expression are intrinsically connected, and that\nthe expression is indistinguishable from the idea, copyright\nprotection cannot be granted. Applying this doctrine courts\nhave refused to protect the expression of an idea that can be\nexpressed only one manner, or in a very restricted manner,\nbecause doing so would confer monopoly on the idea itself. The\ndecision in Herbert Rosenthal Jewelry\nCorporation v. Kalpakian, 446 F.2d 738 (1971) is illustrative in\nthis regard. In that case the plaintiffs sued the defendants asking\nthem to refrain from manufacturing bee shaped jewel pins. The\nCourt held that the jewel shaped bee pin was a an idea that\nanyone  was  free  to  copy,  the  expression  of  which  could  be\npossible only in a few ways, therefore, no copyright could subsist\nin it.\u201d\n[Emphasis supplied]\n\n&#8211; In Akuate Internet Services v. Star India\n## 35\n, Star India was granted\nexclusive broadcasting rights for a cricket match organised by BCCI including\nmobile  distribution  rights.  The  defendants\/appellants  were  disseminating\nmatch information\/score alerts without obtaining a license from the Star India\n(plaintiff\/respondent). In the said case, the plaintiff was seeking copyright\nprotection  over  facts  that  were  available  in  public  domain. The  Division\nBench of this Court held that there cannot be any copyright protection over\nfacts and information that is available in public.\n&#8211; From a reading of the aforesaid judicial precedents, it is clear that there\ncannot be any copyright in facts. Therefore, in the context of news, copyright\nwould subsist only in the form and manner of expression of news and not on\nthe underlying facts. What has to be seen is whether the expression used by\n\n## 35\n2013 SCC OnLine Del 3344.\n\n\n\nCS(COMM) 1028\/2024     Page 50 of 135\n\nOpen AI in its ChatGPT responses is materially similar to the expression used\nby ANI in its news articles resulting in reproduction and making the work\navailable  to  the  public  in  terms  of  Sections  14(a)(i)  and  14(a)(iii)  of  the\nCopyright Act. ANI cannot claim any copyright in the facts underlying the\nnews articles.\n&#8211; ANI  has  given  nine  instances  of  substantial  reproduction  of ANI\u2019s\nworks by Open AI (paragraph 33.1 to 33.9 of the plaint). ANI claims that the\naforesaid nine instances have been given only as an illustration and there are\nmany more instances of copyright infringement by Open AI. All the nine\ninstances given in the plaint are in respect of news\/interviews which were\noriginally published by ANI.\n&#8211; During oral submissions, ANI had highlighted the similarities between\nthe works of ANI and the response produced by ChatGPT in paragraph 33.7\nof the plaint, which dealt with an interview of the mother of Neeraj Chopra\nwho had secured a silver medal at the Olympics, 2024 in Men\u2019s Javelin Throw.\nAdmittedly, the said interview was given in Hindi and was translated into\nEnglish by ANI for reporting. A comparison given by ANI in paragraph 33.7\nof the plaint is set out below:\n\n\n\nCS(COMM) 1028\/2024     Page 51 of 135\n\n\n\n*The colour highlighting has been done for ease of comparison\n&#8211; As per ANI, the response produced by ChatGPT shows that ChatGPT\n\n\n\nCS(COMM) 1028\/2024     Page 52 of 135\n\nhas produced exact or nearly exact copy of ANI\u2019s news article. The original\ninterview and responses were in Hindi which were translated into English by\nANI. Open AI has also infringed ANI\u2019s exclusive rights to translate the works\nfor which ANI invested significant time and effort.\n&#8211; What emerges from a reading of the aforesaid extracts is that ChatGPT\ngenerated this output in response to two different prompts. The first prompt\nwas:\n\u201cPrompt: What did Neeraj Chopra\u2019s mother tell the media about the\nPakistani gold medallist?\u201d\n\n&#8211; A comparison of the response given by ChatGPT to the first prompt as\nhighlighted above (in blue) with what was reported by ANI as highlighted\nabove  (in  orange)  shows  that  the  response  given  by  ChatGPT  was  not  a\nsubstantial  reproduction  or  nearly  exact  copy,  though  the  essence  of  the\nresponse is similar as it based on facts i.e. a particular interview of Neeraj\nChopra\u2019s mother. ChatGPT has added its own commentary to the news article\npublished by ANI, which bears its own expression and is quite dissimilar to\nthe expression used in ANI\u2019s article.\n&#8211; Realising that the response was not a substantial reproduction, ANI\ngave another prompt in the following terms:\n\u201cPrompt : Please try to tell me exactly what she said\u201d\n&#8211; To be noted, in the second prompt, ANI used the word \u2018exactly\u2019, which\nclearly shows that ANI was seeking to elicit a response from ChatGPT in\nexact  terms  i.e.  an  adversarial  prompt  (\u201ccarefully  designed  inputs  that\n\n\n\nCS(COMM) 1028\/2024     Page 53 of 135\n\nmanipulate model outputs\u201d\n## 36\n). In other words, deliberately a prompt is being\nput in a manner so as to elicit a specific response to show that ChatGPT has\nsubstantially reproduced ANI\u2019s works.\n&#8211; Upon the said prompt, ChatGPT came out with a quote of what was\nstated by Neeraj Chopra\u2019s mother as highlighted above (in orange).  The only\nsimilarity which emerges from the aforesaid comparison is relating to a part\nof  a  quote  from  Neeraj  Chopra\u2019s  mother in  the  interview.  However,\nChatGPT\u2019s response has given its own context to the aforesaid quotes from\nNeeraj Chopra\u2019s mother\u2019s interview. In addition, a subsequent paragraph was\nadded (as highlighted in green) completely on its own by ChatGPT to explain\nthe quoted text which was not a part of ANI\u2019s news report.\n&#8211; In  the  present  case, ANI  has  only  demonstrated  that  Open AI  has\nextracted a quote from an interview accompanied by its own commentary.\nAccording to Section 17(cc)\n## 37\nof the Copyright Act, the author of an address\nor speech made in public is the person who made it. The news article gives\nthe quote  of  what  was  said by the  Neeraj  Chopra\u2019s  mother  in  the  video.\nApplying the underlying principle of Section 17(cc) of the Copyright Act to\nthe facts of the present case, in the prima facie view of this Court, Neeraj\n\n## 36\nAhmet  Emre  Erg\u00fcn,  Aytu\u011f  Onan,  Adversarial  Prompt  Detection  in  Large  Language  Models:  A\nClassification-Driven Approach, Computers, Materials and Continua, Volume 83, Issue 3, 2025, Pages 4855-\n4877 ISSN 1546-2218, https:\/\/doi.org\/10.32604\/cmc.2025.063826.\n(https:\/\/www.sciencedirect.com\/science\/article\/pii\/S1546221825004898).\n## 37\n&#8211; First owner of copyright.\u2014Subject to the provisions of this Act, the author of a work shall be the first\nowner of the copyright therein:\nProvided that\u2014\n## &#8230;     &#8230;                &#8230;\n(cc) in the case of any address or speech delivered in public, the person who has delivered such address or\nspeech or if such person has delivered such address or speech on behalf of any other person, such other\nperson shall be the first owner of the copyright therein notwithstanding that the person who delivers such\naddress or speech, or, as the case may be, the person on whose behalf such address or speech is\ndelivered, is employed by any other person who arranges such address or speech or on whose behalf\nor premises such address or speech is delivered;\n\n\n\nCS(COMM) 1028\/2024     Page 54 of 135\n\nChopra\u2019s mother would be first owner of the copyright in the said quotation.\nSimilarly, in terms of Section 14(a)(vii) of the Copyright Act, the right to\ntranslate  would  also  vest  with  Neeraj  Chopra\u2019s  mother.  There  is  nothing\nproduced by ANI to show that Neeraj Chopra\u2019s mother has assigned this\nownership in favour of ANI. Therefore, in the prima facie opinion of this\nCourt, the reproduction of quotes of Neeraj Chopra\u2019s mother by ChatGPT in\nits responses would not amount to infringement of ANI\u2019s copyright.\n&#8211; Pertinently, the original article of ANI (page 54, document no.11 in the\nplaintiff\u2019s documents filed along with the plaint) with which the comparison\nhas been made, was much longer than the text which was quoted in paragraph\n33.7 of the plaint for the purposes of comparison. The response given by\nChatGPT does not reproduce the remaining text which is a part of the original\narticle.\n&#8211; In R.G. Anand (supra), the Supreme Court has held that if a comparison\nhas to be made between two works it should be made for the works as a whole\nand not selected parts of the works.\n&#8211; In this regard, Open AI has relied upon the judgment of Bombay High\nCourt in Leo Burnett (supra). In the said judgment, the Bombay High Court\nheld that a work in which the plaintiff\u2019s claim copyright has to be compared\nas a whole and not in parts in order to determine substantial copying. It is not\npermissible to split or dissect the plaintiff\u2019s work into parts and then compare\nthe  same  with  the  allegedly  infringing  work.  Paragraph  26 of  the  said\njudgment is set out below:\n\u201c26. The issue then is whether if the whole of the respective works is\ncompared and not only parts, factually or materially as it stands, is\nthere copying or substantial copying. I am not at this stage also not\nconsidering whether the \u201cwork\u201d as an episode or all the episodes which\n\n\n\nCS(COMM) 1028\/2024     Page 55 of 135\n\nhave to be considered in totality. Even for arguments sake if an episode\nis considered as a \u2018work\u2019, still what is to be compared is the whole\nwork  as  a  single  entity.  It  is  not  permissible  to  split  or\ncompartmentalize or dissect the plaintiffs\u2019 work into parts, then seek\nto compare such parts against the alleged infringed work. What is to\nbe seen is the entirety of each work. It may be made clear that I am\nnot considering an issue whether a particular part of the film or a\nscene in the film which may have a striking resemblance, if copied,\nwould amount to a substantial copying of the film. For this purpose\nreference may be made to the decision in (Francis Day and Hunter\nLimited v. Twentieth Century Fox Corporation Limited), 1940 A.C. 112\n(Privy  Council);  (Ladbroke  (Football)  Limited  v.  William  Hill\n(Football) Limited), 1964 (1) W.L.R. 273; (Merchandising Corporation\nof America Inc. v. Harpbond Limited), 1983 F.S.R. 32; R.G. Anand v.\nDelux Films supra and (Fritco Lay Limited v. Uncle Chips Private\nLimited), 2000 P.T.C. 341. Applying the tests on the material as it stands\nthere is no substantial copying on facts.\u201d\n[Emphasis supplied]\n\n&#8211; The aforesaid findings have been followed by Coordinate Bench of this\nCourt in Bikramjeet Singh (supra), wherein it was observed that rival works\nhave to be compared as a whole and as a single entity. It is not possible to split\nor dissect the works into parts and then compare them.\n&#8211; During the course of oral submissions, ANI has not addressed on any\nof the other illustrations in paragraph 33 of the plaint. Hence, I do not propose\nto examine the same. Counsel on behalf of Open AI during the course of\nsubmissions has handed over a note to highlight that in none of the instances\ncited by ANI in paragraph 33 of the plaint, ChatGPT has given responses\nidentical or substantially similar to ANI\u2019s articles.\n111.1. The said note has been taken on record.\n111.2. The contents of the said note have not been rebutted on behalf of ANI.\n&#8211; In paragraph 33.9 of the plaint, ANI has stated that on a specific prompt\n\n\n\nCS(COMM) 1028\/2024     Page 56 of 135\n\nabout the latest news published by ANI, ChatGPT has provided updates from\nANI\u2019s website. For the sake of convenience, the screenshot of the specific\nprompt  which  has  been  given  in  paragraph  33.9  of  the  plaint  has  been\ntranscribed below:\n\u201cPrompt &#8211; \u201cGive me the latest news updates from ANI News.in\n\nResponse &#8211; Here are some of the latest news updates from ANI News:\n\n&#8211; In a legal development, AAP leader Durgesh Pathak has been\ngranted bail by the Rouse Avenue court (ANI News).\n&#8211; Indian stock markets saw a positive opening, buoyed by a global\nrally after U.S. inflation data indicated easing in August (ANI\n## News).\n&#8211; The Union Cabinet has approved Phase 3 of  the Bangalore\nMetro Rail Project, which will include two elevated corridors\ncovering 44.65 km with 31 stations (ANI News).\n&#8211; Wrestler Bajrang Punia has moved a petition to the Delhi High\nCourt, leading to a notice being issued to the National Anti-\ndoping Agency (ANI News).\u201d\nFor more detailed updates, you can visit ANI News.\u201d\n\n&#8211; ANI has also filed the corresponding news stories from its website\n(pages 32 to 46 in the documents filed by the plaintiff). The table below gives\na comparison of the title of ANI\u2019s articles and the news updates generated by\nChatGPT based on the prompt set out above.\n## TITLE(S) OF ARTICLE(S) PUBLISHED\n## BY ANI\n## NEWS  UPDATES  GENERATED  BY\n## CHATGPT\nExcise  Police  case:  Delhi  Court\ngrants  bail  to  Durgesh  Pathak,\nArvind Kejriwal\u2019s custody extended\ntill September 25\nIn a legal development, AAP leader\nDurgesh  Pathak  has  been  granted\nbail  by  the  Rouse  Avenue Court\n(ANI News).\n\n\n\nCS(COMM) 1028\/2024     Page 57 of 135\n\nIndian indices opened in green after\nease in US inflation\nIndian stock markets saw a positive\nopening,  buoyed  by  a  global  rally\nafter  U.S.  inflation  data  indicated\neasing in August (ANI News).\nCabinet  approves  two  corridors  of\n## Bangalore Metro Rail Project Phase-\n3  project  for  44.65  km  with  31\nstations\nThe  Union  Cabinet  has  approved\nPhase 3 of the Bangalore Metro Rail\nProject,  which  will  include  two\nelevated  corridors  covering  44.65\nkm with 31 stations (ANI News).\nDelhi HC issues notice to NADA on\nBajrang Punia\u2019s plea challenging his\nsuspension\nWrestler Bajrang Punia has moved a\npetition  to  the  Delhi  High Court\nleading to a notice being issued to\nthe  National Anti-Doping  Agency\n(ANI News).\n\n&#8211; A comparison above shows that even the titles in the response generated\nby ChatGPT are substantially different from the titles of the articles on ANI\u2019s\nwebsite.\n&#8211; Applying the test for infringement laid down in R.G. Anand (supra), in\nmy prima  facie view,  when  the  articles  of  ANI  are  compared  with  the\nresponses generated by ChatGPT as a whole, there is no substantial similarity.\n&#8211; Now, I shall proceed to deal with the judgments relied upon by ANI.\n&#8211; In GEMA v. Open  AI (supra),  the  plaintiff  claimed  copyright\ninfringement  in  respect  of  German  song  lyrics  that  were  memorised  and\nreproduced by the LLMs underlying ChatGPT. The Munich Regional Court\nobserved that the plaintiff therein had expressly excluded the \u2018online search\nfunction\u2019 and in those circumstances a finding of fact was given by the Court\nthat memorisation had occurred as exact song lyrics were reproduced on the\nbasis of non-adversarial prompts. The defendant therein (Open AI) was held\nto be liable for generating outputs that are identical\/substantially similar to the\n\n\n\nCS(COMM) 1028\/2024     Page 58 of 135\n\ncopyrighted lyrics.\n&#8211; In GEMA v. Open AI (supra), the Court found that there were repeated\nverbatim reproduction of lyrics by ChatGPT despite non-adversarial prompts.\nIn that context, the Court held that there was memorisation.\n&#8211; As noted above, in the present case, the training was completed before\nthe illustrative articles in the plaint were published by ANI and therefore, there\ncannot be any question of memorization. Secondly, the Munich Court noted\nthat the responses generated were based on non-adversarial prompts given by\nthe user. In the present case, ANI has given fairly detailed, repeated prompts\nand has even used the term \u2018exactly\u2019, to extract copyrighted contents from the\nLLM  models  underlying  ChatGPT,  which  cannot  be  termed  as  non-\nadversarial. Even after giving such adversarial prompts, ANI was not able to\neke out a response which can be termed as \u2018substantial\u2019 reproduction of its\ncopyrighted material.\n&#8211; In Associated Press v. Meltwater (supra), the defendant-Meltwater was\na  news-monitoring  agency  which  scraped  the  articles  published  by  the\nplaintiff-Associated Press using web crawlers and provided excerpts from\nthese articles to its subscribers through emails or through online account on\nMeltwater\u2019s website. This was not a case of data being scraped by an LLM.\nThe Court in Associated Press v. Meltwater (supra) specifically held that\nthere were 33 articles which were reproduced verbatim by the defendant. In\nthe present case, as noted above, there is no substantial reproduction of ANI\u2019s\nworks.\n&#8211; In Cohere (supra), the plaintiffs, publishers of news and magazines in\nthe  USA,  claimed  that  the  defendant\u2019s  LLM  Models  were  using  the\ncopyrighted   works   of   the   plaintiffs   for   training   and   were\n\n\n\nCS(COMM) 1028\/2024     Page 59 of 135\n\nreproducing\/delivering  verbatim  copies  of  the  plaintiffs\u2019  works  using  the\nRAG function. This judgment was in the context of motion to dismiss, which\nis similar to Order VII Rule 11 of the CPC in India. In the said case, the\nplaintiffs had given 75 instances of copying, out of which, 50 were alleged to\nbe verbatim. Hence, the motion to dismiss was denied.\n&#8211; ANI has also placed reliance on the judgment of the United States Court\nof  Appeal  for  the  Fifth  Circuit  in Positive  Black  Talk v. Cash  Money\n## Records\n## 38\nin support of its submission that copyright infringement is made\nout if there is \u2018probative similarity\u2019 between the plaintiff\u2019s work and the work\nof  the  defendant.  However,  this  decision  was  in  the  context of  copyright\ninfringement in respect of a song and hence, the observations made by the\nCourt were  specific  to  the  said  factual  background. The  application  of  a\nstandard for determining whether there is copyright infringement in respect\nof a song would be quite different from its application in the context of news.\nThe creativity involved in writing lyrics of a song would ordinarily not be\npresent to the same extent in a news article as the fundamental purpose of a\nnews  article  is  to  report  events  that  have  actually  occurred.  Therefore,\nthreshold of establishing substantial similarity in expression would be higher\nin the case of news. Therefore, this judgment would not be of much assistance\nto ANI in the present case.\n&#8211; The  judgments  cited  on  behalf  of  ANI  and  the  intervenor-DNPA,\nInfopaq International v. Danke  Dagblades  Forening\n## 39\nand Newspaper\nLicensing Agency v. Meltwater Holding\n## 40\n, would have little relevance in the\n\n## 38\n## 394 F.3d 357, 368 (5th Cir. 2004).\n## 39\n## ECLI:EU:C:2009:465.\n## 40\n[2010] EWHC 3099 (Ch).\n\n\n\nCS(COMM) 1028\/2024     Page 60 of 135\n\npresent case. In both the aforesaid judgments, it was held that the defendants\nwere  providing  verbatim  extracts  of  the  plaintiff\u2019s  works.  In Meltwater\n(supra), it was held that the defendants have produced extracts and headlines\nof news articles of the plaintiff therein and had accordingly infringed the\ncopyright in the news articles. Same was the position in Infopaq (supra)\nwhere the defendant had copied extracts of news articles of the plaintiff. In\nthe present case, as noted above, there is no substantial reproduction of ANI\u2019s\nworks.\n## FINDINGS\n&#8211; As highlighted above, the illustrations given in the plaint are post the\ntraining of Open AI\u2019s LLMs and a case for memorization of ANI\u2019s works on\nthe basis of the said illustrations cannot be made out. Therefore, at this prima\nfacie stage,  the  contention  of ANI  that  Open AI  permanently  stores  the\ntraining data in order to memorize and regurgitate ANI\u2019s works cannot be\naccepted. At best, these are disputed questions which can only be determined\nduring trial upon parties leading evidence.\n&#8211; Based on discussion above it cannot be concluded that the responses\nproduced by ChatGPT are a substantial reproduction of the news articles of\nANI. Hence, at a prima facie stage, ANI has failed to make out a case of\ncopyright infringement based on the responses generated by ChatGPT i.e. the\noutput claim.\n&#8211; As noted earlier, Issue nos. 1 and 3 are intertwined and have to be\nconsidered together. Hence, I now proceed to deal with Issue nos. 1 and 3.\n\n\n\n\nCS(COMM) 1028\/2024     Page 61 of 135\n\nIssue No.1 Whether the storage by the defendants of plaintiff\u2019s data\n(which is in the nature of news and is claimed to be\nprotected under the Copyright Act, 1957) for training its\nsoftware i.e. ChatGPT, would amount to infringement of\nplaintiff\u2019s copyright\n\n## SUBMISSIONS\nSubmissions on behalf of the Plaintiff (ANI)\n&#8211; Mr. Sidhant  Kumar,  counsel  appearing  on  behalf  of  ANI  made  the\nfollowing submissions:\n127.1. Open AI collects and stores the \u2018raw data\u2019 as a repository of ANI\u2019s\nworks in their original form. The \u2018raw data\u2019 is processed further and filtered.\nThe process of filtration is impossible without storage of the copyrighted\nworks of ANI.\n127.2. Tokenisation of raw data is nothing but numerical representation of\nraw  data  which  belongs  to  ANI.  A  token  may  constitute  an  individual\ncharacter, whole words or even parts of words. The LLM decodes these tokens\nand reconstructs the raw data to generate responses.\n127.3. In the second step, the tokens are encoded into vectors, which are\nthereafter stored in the memory of the LLM to create a vector database. Even\nif it is assumed that the vector database is an adaptation, such reproduction in\nadapted form also results in infringement of copyrighted works in terms of\nSection 2(a)(v) of the Copyright Act.\n127.4. Training  of  LLM  inherently  involves  storage  and  reproduction  of\noriginal works. The entire processes undertaken by Open AI including storage\n\n\n\nCS(COMM) 1028\/2024     Page 62 of 135\n\nof data, tokenisation, vectorisation amounts to copyright violation in terms of\nSection 51 read with Section 14 of the Copyright Act.\n127.5. What Open AI uses in training is only the expression used by ANI and\nnot the actual facts that occurred. Therefore, training occurs in respect of\nexpression of language.\n127.6. Once the training data is fed into Open AI\u2019s LLM during the training\nprocess, it cannot be removed and continues to be reflected in the outputs\ngenerated by the LLM.\n\nIntervenors supporting ANI\ni. Submission  on  behalf  of  Intervenor \u2013 Indian  Music  Industry  (I.A.\n## 4027\/2025)\n&#8211; Mr. C.M. Lall, senior counsel appearing on behalf of intervenors in I.A.\n4027\/2025 made the following submissions:\n128.1. The right to reproduce a work under Section 14(a)(i) of the Copyright\nAct  includes  the  right  to  store  the  work.  Therefore,  scraping,  collection,\ncoalition  and  processing  of  data  for  training  purposes  constitutes\ninfringement.\n128.2. Just  because  copyrighted  works  are  available  on  online\nresources\/websites   would   not   mean   that   these   works   can   be\ncopied\/reproduced without a license or authorization.\n128.3. Regardless of the purpose of the activity, storage of substantial part of\ncopyrighted work amounts to infringement.\n\n\n\n\n\nCS(COMM) 1028\/2024     Page 63 of 135\n\nii. Written Submissions filed by Mr. Ankit Sahni, Advocate on behalf of\nIntervenor \u2013 Federation of Indian Publishers (I.A. 470\/2025)\n&#8211; Open AI  admits  that  collection  and  compilation  of  raw  data  into\ntraining data sets involves creation of copies of the training data. The various\nactivities carried out by Open AI makes it clear that Open AI not only collects\ndata, it also stores the same for training its LLM.\n129.1. Open AI\u2019s act of copying\/storing copyrighted works for purposes of\ntext and data mining\/text and data analysis amounts to infringement since\nthere is no specific provision in the Copyright Act creating an exception for\nthe same.\n\niii. Submissions  on  behalf  of  Intervenors \u2013 Digital  News  Publishers\nAssociation &#038; Ors. (I.A. 2199\/2025)\n&#8211; Mr. Rajshekhar Rao, senior counsel appearing on behalf of intervenors\nin I.A. 2199\/2025 made the following submissions:\n130.1. By using  the  words  \u201cany  medium\u201d  and  \u201cany  material  form\u201d,\nParliament has made it clear that Section 14(a)(i) of the Copyright Act must\nnot be diluted. Merely because protected content is not perceivable by naked\neye or meant for human consumption would not take it outside the ambit of\nSection 14 of the Copyright Act. There is no requirement in Indian copyright\nlaw for the copyrighted works to be stored in a tangible medium. Therefore,\nstorage of literary works in digital formats and their reproduction in the shape\nof  vectors\/tokens  would  fall  within  the  sweep  of  the  expression\n\u201creproduction\/storage\u201d in \u201cany medium\u201d or in \u201cany material form\u201d.\n130.2. The  scheme  of  the  Copyright  Act does  not  distinguish  between\nexpressive and non-expressive uses. The very same argument raised by the\n\n\n\nCS(COMM) 1028\/2024     Page 64 of 135\n\ndefendant in Richard Kadrey v. Meta Platforms\n## 41\nwas rejected by the United\nStates District Court. Even in Bartz v. Anthropic (supra), it was noted that\n\u201cliterary works were chosen for their expressive qualities in building a central\nlibrary and then in training specific LLMs\u201d.\n130.3. Even the intermediate step of introducing and storing data for training\ncannot be said to be non-infringing as substantial portions of copyrighted\nworks are extracted. Once again, this view finds support from the judgment\nin Richard Kadrey v. Meta Platforms (supra).\n\nSubmissions of the Defendant (OPEN AI)\n&#8211; Mr. Amit Sibal, senior counsel appearing on behalf of Open AI made\nthe following submissions:\n131.1. It is an admitted position that storage of data takes place only for the\npurposes  of  training.  Creation  of  copy  during  the  training  process  is  an\nintermediate step. Once the training process is over, an LLM does not have\naccess to data.\n131.2. Purposive  interpretation should  be  given  by  the Courts  to  the\nprovisions of Copyright Act. Accordingly, it is submitted that Section 14 (a)(i)\nof the Copyright Act should not be interpreted in a literal manner. In this\nregard, reliance is placed on:\ni. Shailesh Dhairyawan v. Mohan Balkrishna Lulla\n## 42\n## ;\nii. State of Maharashtra v. Praful B. Desai\n## 43\n## .\n\n## 41\nKadrey v. Meta Platforms, Inc., 3:23-cv-03417, (N.D. Cal.).\n## 42\n## (2016) 3 SCC 619.\n## 43\n## (2003) 4 SCC 601.\n\n\n\nCS(COMM) 1028\/2024     Page 65 of 135\n\n131.3. Training is not based on ANI\u2019s data alone. ANI\u2019s data along with other\ndata is put in database. ANI\u2019s data would only be a miniscule amount of the\ntotal data on which training takes place.\n131.4. For  the  purposes  of  training,  Open  AI  uses  the  non-expressive\nelements  of  literary  works  of ANI  such  as  grammar,  syntax  or  linguistic\npatterns.\n131.5. The  training  process  is  unconcerned  with  each  work\u2019s  unique\nexpression. The LLMs underlying ChatGPT encode the learnt meanings in the\nform  of  a  \u2018vector  database\u2019  during  the  pre-training  process,  which  is\nequivalent to a dictionary and does not amount to additional copies of raw\ndata. Hence, it is incorrect to state that \u2018vectorization\u2019 is an \u2018adaptation\u2019 or\n\u2018translation\u2019 of ANI\u2019s works.\n131.6. The  storage  is  not  intended  for  human  consumption  or\ncomprehension.  Storage  is  only  an  intermediate  step  to  derive  the  non-\nexpressive element.\n\nIntervenors supporting OPEN AI\ni. Submissions on behalf of Intervenor \u2013 Broadband India Forum (I.A.\n## 10766\/2025)\n&#8211; Mr. Kapil Sibal and Mr. Arvind P. Datar, senior counsel appearing on\nbehalf of the intervenor in I.A. 10766\/2025 made the following submissions:\n132.1. Any  storage  of  copyrighted  works  for  training  LLMs  should  be\nprotected  in  line  with  the  statutory  intent  of  making  and  enabling  new\ntechnologies. The storage of works is incidental i.e. training of LLMs and\nLLMs themselves do not store the works.\n\n\n\n\nCS(COMM) 1028\/2024     Page 66 of 135\n\nii. Submission on behalf of Intervenor \u2013 Flux Labs AI Private Ltd. (I.A.\n## 4615\/2025)\n&#8211; Ms.  Haripriya  Padmanabhan, senior  counsel appearing  on  behalf  of\nintervenor in I.A. 4615\/2025 made the following submissions:\n133.1. The  purpose  of  copyright  law  is  that  the  work  ought  not  to  be\ncirculated in public. Taking this purpose into consideration, the reference to\n\u201cstorage\u201d in Section 14(a)(i) of the Copyright Act has to be in the context of\nreproduction of the work. Mere storage of the work would not amount to\ncopyright infringement.\n\nSubmissions of Professor Arul George Scaria, Amicus Curiae\n&#8211; Prof. Scaria made the following submissions:\n134.1. In the present case, the use of copyrighted works by Open AI is \u2018non-\nexpressive use\u2019, hence, it would not amount to infringement. The training data\nis used by Open AI only to extract vital information such as patterns, trend\nand correlations in phrases and sentences.\n134.2. The Court has to consider whether it is feasible to develop an LLM\nwithout giving access to copyrighted material as the strength of the LLM\ndepends upon the extent to which materials are available for training.\n134.3. The Court must consider if all companies developing LLMs have to\nobtain licenses from every copyright holder, it will cause practical difficulties.\nIt will also hinder the development of LLMs by small domestic entities.\n134.4. The expression \u201cstorage\u201d used in Section 14(a)(i) of the Copyright\nAct is only in the context of \u201creproduction\u201d and not storage in an independent\nmanner.  Therefore,  if  there  is  no  reproduction,  there  cannot  be  any\n\n\n\nCS(COMM) 1028\/2024     Page 67 of 135\n\ninfringement on the basis of storage alone. There is no independent right of\nstorage available to the copyright owner.\n\nSubmissions of Mr. Adarsh Ramanujan, Amicus Curiae\n&#8211; Mr. Adarsh Ramanujan made the following submissions:\n135.1. In terms of Section 14 (a)(i) of the Copyright Act, collection of data\nand  storage  of  the  same  would  amount  to  reproduction. Therefore,\ninfringement occurs when Open AI stores the data of ANI. Even if there is a\none-time storage, it would result in infringement.\n135.2. It is likely that while building its corpus of \u2018raw data\u2019, Open AI has\nused and copied ANI\u2019s publicly available works. Even if stored \u2018raw data\u2019 is\ndeleted after the training is complete, it does not change the fact that storage\nhas occurred in the first place.\n135.3. Reproduction of ANI\u2019s works during the training process either in the\noriginal text form or as a collection of numerical tokens, on a prima facie\nview,  constitutes  copyright  infringement.  Such  temporary  copies  of\ncopyrighted  works  do  not  qualify  as \u201ctransient\u201d or \u201cincidental\u201d copies\nprotected under Section 51(1)(b) of the Copyright Act.\n135.4. Even if the purpose of storage is to mine data, the purpose of storage\nor the intention behind it is not relevant for the purposes of infringement under\nSection 51 of the Copyright Act.\n135.5. The subsequent processes of vectorization\/tokenization raise technical\nissues  and  therefore,  whether  these  acts  amount  to  infringement  can  be\ndecided only in a trial.\n135.6. Whether the training process involves further reproduction of ANI\u2019s\ndata, can also be determined only in a trial.\n\n\n\nCS(COMM) 1028\/2024     Page 68 of 135\n\nIssue No.3 Whether the defendants\u2019 use of plaintiff\u2019s copyrighted\ndata qualifies as \u2018fair use\u2019\n## 44\nin terms of Section 52 of the\n## Copyright Act, 1957.\n\n## SUBMISSIONS\nSubmissions of the Defendant (OPEN AI)\n## 45\n\n&#8211; Mr. Amit Sibal, senior counsel appearing on behalf of Open AI made\nthe following submissions:\n136.1. Section 14 of the Copyright Act has to be read with Section 52 of the\nCopyright  Act  as  it  uses  the  words \u201csubject  to  provisions  of  this  Act\u201d.\nAccordingly, Section 52 is not an exception, but a part of what is infringement\nas defined in Section 14 of the Copyright Act. Therefore, the submission on\nbehalf of ANI that Section 52 should be interpreted in a narrow manner,\ncannot be sustained. On the contrary, Section 52 of the Copyright Act should\nbe liberally construed. Reliance in this regard is made on B.D. Bhandari\n(supra) and Chancellor Masters &#038; Scholars of The University of Oxford v.\n## Narendera Publishing\n## 46\n## .\n136.2. Under Section 2(m) of the Copyright Act, a copy of the work would\nbe \u201cinfringing copy\u201d only if the requirements of Sections 14, 51 and 52 of the\nCopyright  Act  are  fulfilled,  as  Section  2(m)  uses  the  expression  \u201cin\ncontravention of the provisions of this Act\u201d. Reliance in this regard is placed\non B.D. Bhandari (supra) and Narendera Publishing (supra).\n\n## 44\nTo be read as \u2018fair dealing\u2019.\n## 45\nSince this defence was raised on behalf of the defendant, it is deemed appropriate that the submissions of\nthe defendant are recorded first.\n## 46\n2008 SCC OnLine Del 1058.\n\n\n\nCS(COMM) 1028\/2024     Page 69 of 135\n\n136.3. The  Copyright Act  as  originally  enacted  in  1957  used  the  words\n\u201cresearch or private study\u201d in its fair dealing clause. The Notes on clauses for\nCopyright  Amendment  Bill,  1994  (which  was  enacted  as  Copyright\nAmendment Act, 1994) makes it clear that the amendment was proposed in\n1992  to  expand  the \u201cunduly narrow  interpretation of  the words \u201cprivate\nstudy\u201d\u201d. The latest amendment in the Copyright Act, 1957 included the words\n\u201cpersonal use\u201d, in addition to the already existing words \u201cprivate use\u201d and\n## \u201cresearch\u201d.\n136.4. There is no restriction on \u201cprivate use\u201d being only for non-commercial\npurposes. This view is reinforced from a perusal of sub-clauses (ii) and (iii)\nof Section 52 (1)(a) of the Copyright Act, which do not make the distinction\nbetween  commercial  and  non-commercial.  Wherever  the  Copyright  Act\nmakes a distinction between commercial and non-commercial uses, it has\nbeen stipulated therein. Reference in this regard may be made to Section\n52(1)(ad) and Section 32(6)(d) of the Copyright Act. Merely because the use\nby Open AI is commercial, would not make it unfair. Reference in this regard\nis made to the judgment in Super Cassettes Industries v. Hamar Television\n## Network\n## 47\n## .\n136.5. ANI has wrongfully placed reliance on the judgment of Rupendra\nKashyap v. Jiwan  Publishing  House\n## 48\n.  In  the  said  case,  the  defendant\nno.3\/CBSE had outsourced printing of its previous years\u2019 question papers to\nthird parties\/defendants no.1 and 2, hence it was held that the defence under\nSection  52  would  not  be  available  as  it  is  available  only  to  a  defendant\nconducting private study or research and not to a third party. B.D. Bhandari\n\n## 47\n2010 SCC OnLine Del 2086.\n## 48\n1996 SCC OnLine Del 466.\n\n\n\nCS(COMM) 1028\/2024     Page 70 of 135\n\n(supra)  holds  that  even  a body  corporate  is  entitled to the defence under\nSection 52(1)(a) of the Copyright Act.\n136.6. The Explanation to Section 52(1)(a) was also added in 2012. In terms\nof the Explanation to Section 52(1)(a) of the Copyright Act, the expression\n\u201cnot itself being an infringing copy\u201d, is restricted to \u201ccomputer programme\u201d\nand not to the other forms of storage. The Explanation\u2019s language and intent\nis clear that the protection of Section 52(1)(a) extends to works stored in an\nelectronic medium for the purposes mentioned under the said provision. The\nlimiting  language  of  the Explanation  within  the  commas  only  applies  to\ncomputer programmes.\n136.7. If ANI\u2019s argument is to be accepted, no electronic storage of a copy of\na work published in an electronic form (such as an online article which is\nfreely accessible over the internet) would ever be eligible for fair dealing\nprotection under Section 52(1)(a). This would mean that making physical\ncopies (for instance, by hand, a typewriter, or a mechanical medium) of a work\nwould be capable of protection under Section 52(1)(a), while the same act\ndone in an electronic medium would be barred from a fair dealing defence.\n136.8. ANI wrongly contends that the defence under Section 52(1)(a) of the\nCopyright Act would not be available to storage as storage is covered under\nSection 52(1)(n) and 52(1)(p) of the Copyright Act. The aforesaid Sections\ncover specific types of storage and on that basis, it cannot be said that storage\nis not covered under Section 52(1)(a) of the Copyright Act.\n136.9. Unlike ANI\u2019s works, which are meant for human consumption for its\nexpressive elements, the use of works for training models by Open AI is non-\nexpressive.\n\n\n\nCS(COMM) 1028\/2024     Page 71 of 135\n\n136.10. The use of Open AI would amount to \u201cprivate use\u201d as mentioned in\nsub-clause (i) of Section 52(1)(a) of the Copyright Act, as it is only for the\npurposes of training  of  the  Open AI\u2019s  LLM.  If  Section 52(1)(a)(i)  of the\nCopyright Act is interpreted to mean that the exception is available only for\npersonal use, the word \u201cprivate\u201d would be rendered otiose.\n136.11. The word \u201cresearch\u201d must be given a broad and liberal interpretation\nso that it is not limited in any manner to a private context. Reliance in this\nregard  is  placed  on  the  judgment  of  Canadian Court in CCH Canadian\n## (supra).\n136.12. Open AI\u2019s storage and use of training data falls within the ambit of\n\u201cprivate  use\u201d,  including \u201cresearch\u201d,  which  is  protected  under  Section\n52(1)(a)(i)  of  the  Copyright Act  and  does  not  amount  to  infringement  of\ncopyright.\n136.13. Indian Courts have adapted the US four-factor test in India. By various\njudgments of the Indian Court, the US four-factor test has been adapted under\nIndian copyright law for the purposes of Section 52(1)(a) of the Copyright\nAct. The Indian judgments which have relied upon the four-factor test are as\nfollows:\ni. India TV Independent News Service v. Yashraj Films\n## 49\n\nii. Super Cassettes Industries Limited v. Chintamani Rao\n## 50\n\niii. B.D. Bhandari (supra)\niv. Narendera Publishing (supra)\nIndian Courts have also recognized that the most crucial test out of the four-\nfactor test is transformative character of the use. The use of publicly available\n\n## 49\n(2012) SCC OnLine Del 4298.\n## 50\n(2011) SCC OnLine Del 4712.\n\n\n\nCS(COMM) 1028\/2024     Page 72 of 135\n\nworks  forming  part  of  the  training  data  for  training  the  LLMs  is  highly\ntransformative. [Ref. Narendera  Publishing (supra) and B.D. Bhandari\n## (supra)].\n136.14. Open  AI\u2019s  ChatGPT  is  not  a  substitute  for  subscription  to  news\nagencies, such as ANI.\n\nIntervenors supporting OPEN AI\ni. Submissions  on  behalf  of  Intervenor &#8211; IGAP  Project  LLP  (I.A.\n## 4616\/2025)\n&#8211; Mr. Akhil Sibal, senior counsel appearing on behalf of the intervenor\nin I.A. 4616\/2025 made the following submissions:\n137.1. When  originally  enacted,  Section  52(1)(a)(i)  of  the Copyright Act\nlimited  the  purposes  to  \u201cresearch  or  private  study\u201d.  The  provision  was\namended in 1994 to \u201cprivate use including research\u201d.\n137.2. The use of copyrighted data to train LLMs meets both, the purpose\nand the fairness  limitations  as  envisaged  under  Section  52(1)(a)  of  the\n## Copyright Act.\n137.3. Section  52(1)(a)  of  the  Copyright  Act  does  not  provide  that\ncommercial uses would not be covered therein. In contradistinction, Section\n52(1)(k)(ii), Section 52(1)(l) and Section 52(1)(o) of the Copyright Act have\nexcluded commercial uses. Hence, the legislature has deliberately chosen not\nto put an element of commerciality in Section 52(1)(a) of the Copyright Act.\n137.4. Reliance placed by ANI on Jiwan Publishing House (supra) and B.D.\nBhandari (supra) to  submit  that  commerciality  defeats  fair  dealing  is\nmisplaced. Jiwan Publishing House (supra) was a case where a publisher of\ninfringing books claimed that since its customers are using the books for\n\n\n\nCS(COMM) 1028\/2024     Page 73 of 135\n\npersonal  use,  the  publisher\u2019s  use  should  also  be  considered  private  or\npersonal, which was rejected by the Court.\n137.5. B.D. Bhandari (supra) holds that a commercial use is not fair in the\ncontext of Section 52(1)(h) of the Copyright Act and not in the context of\nSection 52(1)(a) of the Copyright Act. There is no general proposition laid\ndown in the aforesaid judgment that commercial use defeats fair dealing.\n137.6. The purposes mentioned in Section 52(1)(a) of the Copyright Act,\nsuch  as  \u201cresearch\u201d,  \u201creview\u201d  and  \u201creporting of current news and current\naffairs\u201d are well recognized as permissible uses of copyright, which are often\ncarried out for commercial purposes.\n137.7. While interpreting terms such as \u201cresearch\u201d, \u201creview\u201d and \u201cprivate\nuse\u201d, the Court should apply the doctrine of updating construction i.e. the\nParliament,  while  enacting  a  law  is  expected  to  anticipate  future\ndevelopments. Accordingly, the terms \u201cresearch\u201d, \u201cprivate use\u201d and \u201creview\u201d\nmay not be limited to acts of human beings alone but to machine learning as\nwell. Therefore, these terms should be interpreted liberally.\n137.8. The term \u201cprivate use\u201d must be contradistinguished from the term\n\u201cpersonal use\u201d. Supreme Court in Academy of General Education, Manipal\nv. B.  Malini  Mallya\n## 51\nhas  adopted  the  expansive  meaning  of  the  word\n\u201cprivate\u201d by holding that a dance performance at an educational institution\ncan  also  be  within  the  meaning  of  \u201cprivate use\u201d  for  research,  review  or\ncriticism. \u2018Pre-training dataset\u2019 obtained by LLMs is private in nature i.e.\naccessible only to LLM models themselves. In other words, this data set is not\n\n## 51\n## (2009) 4 SCC 256.\n\n\n\nCS(COMM) 1028\/2024     Page 74 of 135\n\npublicly  available  for  either  access  or  download.  Therefore, \u2018use\u2019 of  any\ncopyrighted material is totally private.\n137.9. In B.D.  Bhandari (supra),  the  Division  Bench  adopted  the\ntransformative test adopted by the American Courts for determining what\nconstitutes fair use. The Court considered whether contribution in writing a\n\u2018guide book\u2019 is such that a nature of guide book is different than the textbook.\nIt  was  held  that  if  the  purpose  served  by  the  guide  book  is  substantially\ndifferent from the purpose served by the text book, it would be treated as\ntransformative.\n137.10. In The Chancellor, Masters &#038; Scholars of University of Oxford v.\nRameshwari Photocopy Services (DB)\n## 52\n, it was held that the purposes of use\nwould determine whether a particular use qualifies as being fair. The fairness\nis determined by whether the extent is justified by its purpose. In the context\nof reproduction of literary works for creating course packs, the Court held that\ncopying of entire articles was justified for the purposes of education.\n137.11. The DNPA and IMI have wrongfully relied upon Bartz v. Anthropic\n(supra) to submit that the fair use defence for training is unavailable if the first\ncopy is not authorized by the copyright holder. In Bartz v. Anthropic (supra),\nthe defendant had used \u2018pirated books\u2019 for creating a permanent general-\npurpose library, which was held to be not fair use. In the present case, Open\nAI does not create any library and the data is used only for training purpose.\n137.12. The US Courts in Bartz v. Anthropic (supra) and Kadrey v. Meta\nPlatforms (supra) have come to the conclusion that the use of copyrighted\nmaterials for training LLM is transformative. In Bartz v. Anthropic (supra),\n\n## 52\n2016 SCC OnLine Del 6229.\n\n\n\nCS(COMM) 1028\/2024     Page 75 of 135\n\nthe Court drew parallels between the training process and a human being\nreading books and generating new works. It was observed that the entire\npurpose of training was not to replicate or supplant the original words on\nwhich the LLM was trained but to create something different.\n137.13. Since the purpose and character of the use of ANI\u2019s work is only for\ntraining the data sets along with millions of other works being used for the\nsame purpose, the same would amount to transformative purpose. Further,\nsince the format of storage on the input side is totally different from that of\nthe ANI\u2019s data, i.e. the data stored in a machine-readable algorithm which\ncannot be read by humans, the same would amount to transformative use.\n137.14. In  the  present  case,  the  objective  of  ANI  and  Open  AI  are\nfundamentally different. ANI is a news agency which is mainly focused on\ncreating, verifying and distributing news which involved a process of fact-\ngathering, investigative journalism and contextualization of stories. On the\nother  hand,  LLMs  serve  as  tools  such  as  content  creation,  researching,\nbrainstorming  and  idea  generation,  language  translation,  learning  and\neducation,  creative  writing,  summarizing  or  paraphrasing.  Therefore,\nactivities of Open AI cannot substitute the activities carried out by ANI. No\nevidence has been provided by ANI that Open AI\u2019s use of AI generative use\nhad led to a decrease in their market share.\n\nSubmissions of the Plaintiff (ANI)\n&#8211; Mr.  Sidhant  Kumar,  counsel  appearing  on  behalf  of  ANI  made  the\nfollowing submissions:\n138.1. Open AI\u2019s actions do not fall under \u201cprivate or personal use, including\nresearch\u201d as it is used purely for commercial purposes and for making profit.\n\n\n\nCS(COMM) 1028\/2024     Page 76 of 135\n\nIt is settled law that the defence under Section 52(1)(a)(i) is unavailable to\nperson who uses the copyrighted works for commercial exploitation [Ref.\nJiwan Publishing House (supra)]. Consequently, Open AI\u2019s claim of use for\npersonal and private research is also unsustainable as it uses ANI\u2019s works for\na manifestly commercial purpose, that is, to earn revenues from enhanced\nfunctioning of the ChatGPT model.\n138.2. The contention of Open AI that the condition of \u201cnon-infringing copy\u201d\nis  only  for  a  computer  programme  is  misconceived.  The  Explanation\nappended  to  Section  52(1)(a)  of  the  Copyright Act  extends  the  specified\nexemption to an electronically stored copy of a work subject to the condition\nthat the work itself is lawfully stored. The Explanation is centred around the\nterm \u201cany work\u201d. Consequently, the condition that the work must not \u201citself\u201d\nbe an infringing copy, has to be a reference to the \u201cwork\u201d itself and not to a\ncomputer programme that may be incidentally involved. The word \u201citself\u201d\nemphasizes that the concern is about the copy stored. It does not refer to some\nother copy, not the programme in the abstract, but the particular copy. This is\nsignificant because it echoes the concerns about the distinction between the\nlawfulness of the act (storage for fair dealing) and the lawfulness of the stored\ncopy.\n138.3. The opening line of Section 52(1)(a) expressly excludes computer\nprogrammes from its ambit. It is settled law that Explanation cannot enlarge\nthe scope of the parent provision. Consequently, the Explanation cannot be\nread to provide a fair dealing right with respect to computer programmes that\nthe parent clause itself has denied.\n\n\n\nCS(COMM) 1028\/2024     Page 77 of 135\n\n138.4. The  Explanation  to  Section  52(1)(a)  of  the  Copyright  Act  only\nprovides protection to incidental storage of non-infringing copies. Therefore,\npermanent storage of ANI\u2019s work would amount to infringement of copyright.\n138.5. Open AI\u2019s use of ANI\u2019s works would not qualify as \u201cresearch\u201d under\nSection 52(1)(a)(i) of the Copyright Act. Open AI just copies ANI\u2019s creative\nexpression and generates responses without any creative input from Open AI.\n138.6. Without prejudice to the aforesaid submissions, the four-factor test\nunder the US Law has no application in the Indian copyright Law.\n138.7. Open AI\u2019s claim of transformative use is flawed. The expressive and\nnon-expressive element of raw data are not segregated before feeding the\ninformation in the LLM.\n138.8. Open AI\u2019s ChatGPT works as an alternative to the ANI\u2019s publications\nand diminishes the ANI\u2019s market.\n\nIntervenors supporting ANI\ni. Submission on behalf Intervenor \u2013 Digital News Publishers Association\nand Ors. (I.A. 2199\/2025)\n&#8211; Mr. Rajshekhar Rao, senior counsel appearing on behalf of intervenors\nin I.A. 2199\/2025 made the following submissions:\n139.1. Under Section 52(1)(a) of the Copyright Act, Open AI has to show the\nfollowing to be eligible to claim the defence to copyright infringement:\n(a) The  works  utilized  were  non-infringing  copies  as  mandated\nunder the Explanation to Section 52(1)(a) of the Copyright Act\nthat have been lawfully accessed.\n\n\n\nCS(COMM) 1028\/2024     Page 78 of 135\n\n(b) The use of works was for any one (or more) of the purposes\nenumerated at Section 52(1)(a)(i), (ii) and (iii) of the Copyright\n## Act.\n(c) The use of the works amounted to fair dealing.\n139.2. Storage  of  copyright  materials  which  are  publicly  available  by  a\ncommercial entity for disseminating information to the public can never be\n\u201cprivate or personal use\u201d. Under Section 52(1)(a) of the Copyright Act, Open\nAI\u2019s LLM is only available to an individual researcher or student, this cannot\nbe claimed by a commercial organization. Since the elemental purpose is\ndissemination  of  commercial  platform  to  subscribers  and  users,  the  same\ncannot amount to \u201cprivate use\u201d.\n139.3. Data  sets  containing  copyrighted  information  are  accessed  and\nextracted unlawfully from sources without a license. Therefore, the defence\nof fair dealing cannot be invoked.\n139.4. As per Indian judgments, the American four-factor test has only been\napplied as a guiding principle to determine fairness of a dealing and only has\na persuasive value. There are significant structural and textual differences\nbetween the Indian concept of fair dealing and American test of fair use. To\nreiterate, right to \u2018transform\u2019 a work is explicitly recognized under the US\nCopyright Act, whereas no such recognition exists under the Indian Copyright\n## Act.\n139.5. The  submission  of  Open AI  that  storage  of  works  for  extracting\ninformation in the process of training an AI model is a transformative purpose,\nis misplaced. The US Supreme Court in Andy Warhol v. Lynn Goldsmith\n## 53\n\n\n## 53\n2023 SCC OnLine US SC 19.\n\n\n\nCS(COMM) 1028\/2024     Page 79 of 135\n\nheld that merely because the secondary work adds new expression to the\nsource material, it will not be transformative.\n139.6. The ultimate purpose of training an LLM and that of a news platform\nis the same i.e. to disseminate accurate information to the public. When a\nsecondary work shares a common purpose with the original, work ceases to\nbe transformative. Reliance is placed on the observations made by the US\nDistrict Court in Kadrey v. Meta Platforms (supra), where the Court has noted\nthe unlawful advantage that training an AI model makes of creative expression\nand  the  effect  it  has  on  the  market  share  for  news  and  other  fact-based\nexpression.\n\nii. Submission on behalf of Intervenor \u2013 Indian Music Industry and Ors.\n## (I.A. 4027\/2025)\n&#8211; Mr. C.M. Lall, senior counsel appearing on behalf of intervenors in I.A.\n4027\/2025 made the following submissions:\n140.1. The copies of works used by AI models are obtained from unlawful\nsources\/works used without a license from the copyright owner.\n140.2. Use of any copyright protected content for the purposes of training AI\nmodels does not ex facie amount to fair dealing of the works under Section\n52(1)(a) of the Indian Copyright Act.\n140.3. The Explanation, which was introduced by the 2012 Amendment to\nthe Copyright Act, effectively incorporated an additional condition to Section\n52(1)(a)  that  the  storage  of  any  work  for  the  \u201cpurposes\u201d  under  Section\n51(1)(a) must be of a non-infringing copy of the work.\n140.4. The word \u201cincluding\u201d occurring between \u201cstoring of any work in any\nelectronic  medium  for  the  purposes  mentioned  in  this  clause\u201d  and  \u201cthe\n\n\n\nCS(COMM) 1028\/2024     Page 80 of 135\n\nincidental storage of any computer programme\u201d is the crucial connecting link.\nIt is settled law that the usage of the word \u201cincluding\u201d in statutory provisions\nis conjunctive. The placement of commas does not control the meaning of the\nExplanation to Section 52(1)(a). The opening part of the Explanation also\nrefers to \u201cany work\u201d.\n140.5. The ultimate purpose of the Open AI is not research or private or\npersonal use, but to develop and make available a commercial AI model to its\nsubscribers\/users. The word \u201cprivate\u201d would have to be viewed in the context\nof the word \u201cresearch\u201d. Open AI\u2019s activities are not entitled to any defences\navailable under Section 52 of the Copyright Act.\n140.6. The wholesale reproduction of entire copyrighted content would not\namount to fair dealing. The use of the word \u201cincluding\u201d in Section 52(1)(a) is\nin fact a limitation and not an expansion. The expression \u201cprivate use\u201d thus\neffectively ousts \u2018commercial use\u2019. The use by Open AI does not amount to\n\u201ctransient or incidental storage\u201d which is exempted under Section 52(1)(b)\nof the Copyright Act.\n140.7. The  open-ended  US  Copyright  Law  Doctrine  of \u2018Fair  Use\u2019 is  not\nsanctioned by the statute in India. The limitations under Section 52(1)(a) of\nthe Copyright Act would have to be satisfied in order to invoke the defence\nspecified therein. The concept of \u2018transformative use\u2019 applied by American\nCourts cannot  be  adopted  in  the  Indian  context. The Courts  in  UK  and\nAustralia have desisted from incorporating the \u2018fair use\u2019 principle into their\ndomestic law.\n\n\n\nCS(COMM) 1028\/2024     Page 81 of 135\n\n140.8. The  judgments  of  the  US Courts  in Thomson  Reuters v. Ross\n## Intelligence\n## 54\n, Bartz v. Anthropic (supra) and Kadrey v. Meta Platforms\n(supra), support the view that use of the copyrighted material for training an\nLLM cannot be considered fair use.\n\niii. Written Submissions filed by Mr. Ankit Sahni, Advocate on behalf of the\nIntervenors \u2013 Federation of Indian Publishers (I.A. 470\/2025)\n&#8211; The use of the expression \u201cnamely\u201d in Section 52 makes it clear that\nonly the acts that are specifically mentioned in Section 52 are exempted from\nbeing construed as infringement.\n141.1. The  defence  under  Section  52(1)(a)(i)  of  the  Copyright  Act  is\navailable  to  an  individual  user  and  not  to  a  defendant  who  is  exploiting\ncopyrighted  works  for its  commercial  activities. [Ref. TIPS Industries v.\n## Wynk Music\n## 55\n## ]\n141.2. The de-minimis defence taken by Open AI that the copyrighted work\nused by Open AI forms very miniscule part of the news related work being\nused by Open AI is misplaced as de-minimis is to be seen qua the extent of\nthe work of the copyrighted owner that has been used by the defendant and\nnot qua the entire copyrighted work of the various copyrighted owners being\nused by the defendant.\n141.3. The four-factor test adopted in the US in respect of fair use would not\nbe applicable in India.\n\n\n\n\n## 54\nCase No. 1:20-cv-613-SB.\n## 55\n2019 SCC OnLine Bom 13087.\n\n\n\nCS(COMM) 1028\/2024     Page 82 of 135\n\nSubmissions of Professor Arul George Scaria, Amicus Curiae\n&#8211; Prof. Scaria made the following submissions:\n142.1. The fair dealing analysis requires a two-step assessment:\ni. The first step, whether the use falls under one of the specific\npurpose categories mentioned in Section 52 of the Copyright Act.\nii. The second step, whether the use can be considered as \u201cfair\u201d or\nnot.\n142.2. The  terms  \u201cprivate  use\u201d  and  \u201cresearch\u201d  mentioned  in  Section\n52(1)(a)(i) of the Copyright Act and the words \u201ccriticism or review\u201d in Section\n52 (1)(a)(ii) of the Copyright Act and \u201creporting of current news events\u201d in\nSection 52(1)(a)(iii) of the Copyright Act must be interpreted in a manner\nwhich is consistent with the broader purpose of copyright law i.e. to promote\ncreation and dissemination of new works.\n142.3. As the learning process of the LLM does not involve giving access to\nany  external  human  being  and  such  learning  is  primarily  done  with  the\nobjective of promoting research, such storage would be covered within the\nbroad ambit of Section 52(1)(a)(i) of the Copyright Act.\n142.4. Insofar  as  interplay  between  Section  51  and  Section  52  of  the\nCopyright Act is concerned, the Court should take into account the opening\nwords of Section 52 which provide that the \u201cfollowing acts shall not constitute\nan infringement\u201d. Therefore, Section 52 should be interpreted by a Court in a\nliberal manner. Reference is made to Single Bench judgment in Chancellor,\nMasters &#038; Scholars of the University of Oxford v. Rameshwari Photocopy\n\n\n\nCS(COMM) 1028\/2024     Page 83 of 135\n\n## Services\n## 56\nand  the  judgment  of  the  Supreme Court in B. Malini Mallya\n## 57\n\n## (supra).\n142.5. There is no requirement in Section 52(1)(a) of the Copyright Act for\nOpen AI to acquire lawful access to ANI\u2019s copyrighted works. Wherever a\nstatute requires lawful access the same has been specifically provided in the\nstatute [Ref. Section 52(1)(ab) and Section 52(1)(ad)].\n142.6. Section 52(1)(a) of the Copyright Act specifically excludes computer\nprogramme  as a computer  programme  is  specially  covered  under  other\nprovisions of Section 52(1) of the Copyright Act. The Explanation to Section\n52(1)(a)  is  in  respect  of  incidental  use  of  computer  programme  for  the\npurposes of storage of a work.\n142.7. There is no prohibition for \u2018commercial use\u2019 under Section 52(1)(a)(i)\nof the Copyright Act. Therefore, the same is not a relevant factor for the\npurpose of Section 52(1)(a)(i) of the Copyright Act.\n142.8. Section 29 of the Canadian copyright law also provides a similar fair\ndealing  exception  as  provided  under  the  Indian  Copyright Act.  In CCH\nCanadian (supra),  Canadian  Supreme Court held  that  the  expression\n\u201cresearch\u201d used in Section 52(1)(a)(i) of the Copyright Act should be given a\nbroad and liberal meaning. Reliance in this regard is also placed on judgment\nof the Canadian Supreme Court in Society of Composers, Authors and Music\nPublishers of Canada v. Bell Canada\n## 58\n## .\n142.9. The four-factor test adopted in United States is not applicable in India.\nTherefore, the Courts in India are free to adopt an appropriate test to determine\n\n## 56\n2016 SCC OnLine Del 6713, \u00b6 28.\n## 57\n## \u00b6\u00b6 38-39.\n## 58\n## [2012] 2 S.C.R. 326.\n\n\n\nCS(COMM) 1028\/2024     Page 84 of 135\n\nfair dealing. As is evident from the judgments passed by Indian Courts, there\nis no single approach that the Courts have used to determine fair dealing.\n\nSubmissions of Mr. Ramanujan, Amicus Curiae\n&#8211; Mr. Ramanujan made the following submissions:\n143.1. Section 52(1)(a) of the Copyright Act should not be expanded beyond\nits expressed terms is apparent from the use of the term \u201cnamely\u201d, which\nsuggests that the list which follows is exhaustive.\n143.2. The use of the words \u201cprivate or personal\u201d in Section 52(1)(a)(i) of\nthe  Copyright Act  indicate  that  the  exemption  is  intended  to  apply  at  a\npersonal level to an individual and not to a defendant with a commercial\nmotive. [Ref. TIPS Industries v. Wynk Music (supra)].\n143.3. The four-factor test adopted by the US Courts cannot be adopted in\nIndia in the absence of any statutory framework. The same can be used only\nto supplement and not to supplant the statutory mandate of Section 52 of the\nCopyright Act. The Indian fair dealing exception is purpose limited, whereas\nthe US \u2018fair use\u2019 exception  is purpose agnostic. [Ref. Super Cassettes v.\nHamar Television (supra); Jiwan Publishing House (supra) and Syndicate\nof Press of University of Cambridge v. Kasturi Lal\n## 59\n## ].\n143.4. The judgment of Division Bench in B.D. Bhandari (supra), clearly\nlays down that Section 52 is in the nature of an exception to Section 51 of the\nCopyright Act. Therefore, there should be a two-step analysis \u2013\n(i) whether there is infringement and,\n\n## 59\n2005 SCC OnLine Del 1448.\n\n\n\nCS(COMM) 1028\/2024     Page 85 of 135\n\n(ii) whether the infringing act falls in any of the exceptions provided\nunder Section 52 of the Copyright Act.\n143.5. The expression \u201cprivate use\u201d in Section 52(1)(a)(i) of the Copyright\nAct would not apply to Open AI in the present case. Even though the process\nadopted by Open AI for training is private, the use is not private. As a result\nof the training, the output which is produced is for the use by the public.\n143.6. Reliance placed by Open AI on the judgment in B. Malini Mallya\n(supra) may not be apposite as the observations made in paragraph 38 of the\nsaid judgment that institutions can take benefit of Section 52 of the Copyright\nAct were \u2018sub silentio\u2019 observations.\n143.7. The expressions \u201cprivate\u201d or \u201cpersonal\u201d used in Section 52(1)(a)(i) of\nthe Copyright Act have been used in an overlapping manner and would apply\nonly to use by individuals and not by corporations like Open AI. The judgment\nin Narendera Publishing (supra), was in the context of Section 52(1)(a)(ii)\nand not Section 52(1)(a)(i) of the Copyright Act. Sub-clauses (ii) and (iii) of\nSection 52(1)(a) do not use the word \u201cprivate\u201d or \u201cpersonal\u201d. The Single\nBench applied the transformative test in the context of \u201creview\u201d under Section\n52(1)(a)(ii) of the Copyright Act and not in the context of fair dealing.\n143.8. In India, the test to be applied to determine fair dealing would be\n\u2018substantiality of copying\u2019 and \u2018motive to compete\u2019 as stated in Blackwood v.\nA.N. Parasuraman\n## 60\n, which was cited in Rameshwari Photocopy Services\n(DB) (supra).\n143.9. In B.D. Bhandari (supra), the test that was adopted by the Division\nBench  in  the  context  of  Section 52(1)(h)  was  whether  a  guide  book  is a\n\n## 60\n1958 SCC OnLine Mad 62.\n\n\n\nCS(COMM) 1028\/2024     Page 86 of 135\n\n\u2018derivative work\u2019, and the test of what would amount to derivative work was\nlaid down in Eastern Book Company v. D.B. Modak (supra). In paragraph\n119  of  the  said  judgment,  the  said  test  of  derivative  work  has  also  been\nextended  to  Section  52(1)(a)(ii)  of  the  Copyright  Act.  It  appears  the\nexpression \u2018derivative work\u2019 has been used synonymously with the expression\n\u2018transformative work\u2019. Therefore, both the judgments of the Single Bench and\nthe Division Bench in B.D. Bhandari (supra) do not provide a definitive\nruling that \u2018transformative work\u2019 would apply to Section 52(1)(a)(i) of the\n## Copyright Act.\n143.10. The exception of \u2018transformative use\u2019 is a part of the first factor of the\nfour-factor test, which is applied in the United States of America and the same\nwould not be applicable in India.\n\n## DISCUSSION AND ANALYSIS OF ISSUE NOS. 1 AND 3\n&#8211; Now, this Court shall examine the Issue Nos. 1 and 3, as follows:\nA. Scope of Section 14(a)(i) of the Copyright Act, 1957\n\n&#8211; As observed in Issue No. 2, this Court is of prima facie view that the\ncopyright in respect of original literary works available on the website of ANI\nvests with ANI. Therefore, ANI enjoys certain exclusive rights under Section\n14 of the Copyright Act, which includes the right of \u201creproduction\u201d under\nSection 14(a)(i) of the Copyright Act. The right of reproduction includes the\nright of storing the literary works owned by ANI \u201cin any medium by electronic\nmeans\u201d. For ease of reference, Section 14(a)(i) of the Copyright Act is set out\nbelow:\n\n\n\n\nCS(COMM) 1028\/2024     Page 87 of 135\n\n\u201c14. Meaning of Copyright\u2014\nFor  the  purposes  of  this  Act, \u201ccopyright\u201d means  the exclusive  right\nsubject to the provisions of this Act, to do or authorise the doing of\nany of the following acts in respect of a work or any substantial part\nthereof, namely:\u2014\n(a)  in  the  case  of  a  literary,  dramatic  or  musical work,  not  being  a\ncomputer programme,\u2014\n(i) to  reproduce  the  work  in  any  material  form  including  the\nstoring of it in any medium by electronic means;\u201d\n\n[Emphasis supplied]\n\n&#8211; The original Section 14(1)(a)(i) as enacted on 4\nth\nJune, 1957 only gave\nthe owner of the literary work, the \u201cright to reproduce it in any material form\u201d.\nThe phrase \u201cincluding the storing of it in any medium by electronic means\u201d\nwas incorporated in Section 14(a)(i) through the Copyright (Amendment) Act,\n1994 enacted on 9\nth\nJune, 1994. A comparison between the original text of\nSection 14 (relevant part) and the amended text is tabulated below:\n## THE COPYRIGHT ACT, 1957 [4\nth\n## June\n## 1957]\n## THE COPYRIGHT (AMENDMENT) ACT,\n## 1994 [9\nth\n## June 1994]\n&#8211;  (1)  For  the  purposes  of  this  Act,\n\u201ccopyright\u201d means the exclusive right, by\nvirtue of, and subject to the provisions of,\nthis Act, \u2014\n(a)  in  the  case  of  a  literary,  dramatic  or\nmusical  work,  to  do  and  authorise  the\ndoing of any of the following acts, namely:\n## \u2014\n(i) to  reproduce  the  work  in  any\nmaterial form,\n&#8211;  For  the  purposes  of  this  Act,\n\u201ccopyright\u201d means  the  exclusive  right\nsubject to the provisions of this Act, to do\nor  authorise  the  doing  of  any  of  the\nfollowing acts in respect to a work or any\nsubstantial part thereof, namely:\u2014\n(a)  in  the  case  of  a  literary,  dramatic  or\nmusical  work,  not  being  a  computer\nprogramme,\n(i) reproduce the work in any material\nform including the storing of it in any\nmedium by electronic means;\n\n\n\n\n\n\nCS(COMM) 1028\/2024     Page 88 of 135\n\n&#8211; By way of the said amendment, the right of reproduction was expanded\nto incorporate the digital storage of a literary work. This amendment was\nenacted to keep pace with the evolving technological realities.\n&#8211; The scope and ambit of Section 14(a)(i) of the Copyright Act insofar as\nstorage of a literary work is concerned, is very wide as it uses the expression\n\u201cstoring of it in any medium by electronic means\u201d and the expression \u201cin any\nmaterial form\u201d. In terms of Section 14(a)(i) of the Copyright Act, the storage\nof a copyrighted literary work in any medium by electronic means, would\ntherefore, amount to reproduction of the work.\n&#8211;  Section 14 (a)(i) of the Copyright Act, when it speaks of storage, does\nnot make a distinction between temporary or permanent storage. Similarly,\nunder Section 51\n## 61\nof the Copyright Act, the intention\/purpose of storage is\nnot relevant when determining an infringement claim. Thus, if a literary work\nis stored temporarily or permanently in any medium using electronic means,\nthe same would violate the exclusive right of the copyright holder as per\nSection 14(a)(i) of the Copyright Act and would be an act of infringement\nunder Section 51 of the Copyright Act.\n&#8211; However, it is pertinent to note that Section 14 of the Copyright Act\nuses the expression, \u201csubject to the provisions of this Act\u201d, thus making it clear\n\n## 61\n&#8211; When copyright infringed.\u2014Copyright in a work shall be deemed to be infringed\u2014\n(a) when any person, without a licence granted by the owner of the Copyright or the Registrar of Copyrights\nunder this Act or in contravention of the conditions of a licence so granted or of any condition imposed by a\ncompetent authority under this Act\u2014\n(i) does anything, the exclusive right to do which is by this Act conferred upon the owner of the copyright,\nor\n(ii) permits for profit, any place to be used for the communication of the work to the public where such\ncommunication constitutes an infringement of the copyright in the work, unless he was not aware and had\nno reasonable ground for believing that such communication to the public would be an infringement of\ncopyright; or\n\n\n\n\nCS(COMM) 1028\/2024     Page 89 of 135\n\nthat the exclusive rights therein would be subject to other provisions of the\nCopyright Act. The other provisions would also necessarily include Section\n52 of the Copyright Act which provides for various acts that do not constitute\ninfringement under the Copyright Act. In this regard, reference may be made\nto opening sentence of Section 52 of the Copyright Act:\n\n\u201c52.  Certain  acts  not  to  be  infringement  of  copyright.\u2014(1)  The\nfollowing  acts  shall  not  constitute  an  infringement  of  copyright,\nnamely\u2014\u201d\n\n\nThus, specific acts enlisted under Section 52 of the Copyright Act cannot be\nconsidered to be falling under Section 14 and 51 of the Copyright Act.\n&#8211; Therefore, while analysing the rights of copyright holder under Section\n14 of the Copyright Act, the Court would also have to bear in mind Section\n52 of the Copyright Act as that would make the \u2018act\u2019 which would have been\notherwise infringing under Section 51, to be non-infringing. The expression\n\u201cinfringing copy\u201d as defined under Section 2(m)\n## 62\nof the Copyright Act also\nuses the words \u201cin contravention of the provisions of this Act\u201d, therefore, if\nthe work\/act is in accordance with provisions of Section 52 then it would not\nbe an infringing copy.\n&#8211; This  view  also  finds  support  from  the  observations  made  by  the\nDivision Bench in B.D. Bhandari (supra). In the said decision, the Division\nBench observed that when a defendant raises a defence under Section 52 of\n\n## 62\n&#8211; Interpretation.\u2014In this Act, unless the context otherwise requires, \u2014\n## &#8230;     &#8230;     &#8230;\n(m) \u201cinfringing copy\u201d means,\u2014\n(i) in relation to a literary, dramatic, musical or artistic work, a reproduction thereof otherwise than in the\nform of a cinematographic film;\n## &#8230;     &#8230;     &#8230;\nif such reproduction, copy or sound recording is made or imported in contravention of the provisions of this\n## Act;\n\n\n\nCS(COMM) 1028\/2024     Page 90 of 135\n\nthe Copyright Act, the Courts are expected to give a finding whether the act\nof the defendant falls within the categories enlisted under the said section\nbefore giving a finding of infringement. The relevant paragraphs from B.D.\nBhandari (supra) are set out below:\n\n\u201c38. It is, thus, clear that when a work is in public domain, the natural\nconsequence  is  that  it  does  not  enjoy  copy  right  and  question  of\nviolation  thereof  would  not  arise  at  all  inasmuch  as  anybody  and\neverybody has right to use and\/or copy the same. On the other hand,\nwhen we talk of \u201cfair use\u201d, it would in the context that there is someone\nenjoys copyright in that work, but the user thereof comes within the\ndomain of eventualities provide under Section 52 of the Copyright Act.\nThus,  all  the  cases  of  fair dealing  are  infringements,  but  the  law\nexamines the same based on the principles of \u201cmodicum of creativity\u201d,\naccess  and  affordability  based  on  social,  economic,  educational,\nindustrial, etc. considerations of society.\n&#8211; What follows form the above is that the two concepts, viz., \u2018public\ndomain\u2019 and \u2018fair use\u2019 work in altogether different fields. When the\ndefence is taken that a particular work is in public domain, the question\nfor  consideration  would  arise  as  to  whether  the  appellants  enjoy\ncopyright therein or not. On the other hands, when plea of fair use\nunder Section 52 of the Copyright Act is adopted by the respondents\nwhat is expected from the Court on the application of the provisions\nof Section 52 of the Act is not to examine whether the activity is\ninfringement is not, but to examine whether the conditions stated in a\nparticular clause of Section 52 of the Act invoked by the respondents\nis satisfied or not.\u201d\n[Emphasis supplied]\n\n&#8211; The  Single  Bench  in Rameshwari Photocopy Services (supra) also\nobserved that if an act, which is an exclusive right of the copyright holder, is\nprovided in Section 52 of the Copyright Act, it would not amount to copyright\ninfringement. The  relevant paragraph  of  Rameshwari  Photocopy  Services\n\n\n\n\nCS(COMM) 1028\/2024     Page 91 of 135\n\n(Single Bench) (supra) is set out below:\n\u201c28. Section 51 prescribes that copyright is infringed inter alia when\nany  person  does  anything  exclusive  right  to  do  which  has  been\nconferred by the Act on the owner of copyright. It follows, if there is no\nexclusive right, there is no infringement. Section 52 lists the acts which\ndo not constitute infringement. Thus, even if exclusive right to do\nsomething constitutes copyright, if it finds mention in Section 52,\ndoing thereof will still not constitute infringement and the outcome\nthereof will not be infringing copy within the meaning of Section\n2(m). Section 55 also, entitles the owner of copyright to remedies by\nway of injunction, damages, accounts or otherwise as are conferred by\nlaw for infringement of a right, only when copyright is infringed and\nexcept as otherwise provided by the Copyright Act. Thus unless there is\ninfringement  of  copyright  within  the  meaning  of  the Act,  owner  of\ncopyright is not entitled to sue.\u201d\n[Emphasis supplied]\n\nAn appeal was preferred against the Single Bench judgment. The Division\n## Bench\n## 63\ndid not interfere with the finding of the Single Bench with respect to\ninterpretation of Section 52. In fact, the Division Bench noted that the counsel\nfor  the  parties  were  in  agreement  that  Section  52  is  not  a  proviso  or  an\nexception to Section 51 of the Copyright Act.\n&#8211; Therefore, in my opinion, Issue No.1, as framed by this Court dealing\nwith the storage of ANI\u2019s copyrighted works by Open AI and Issue No.3 on\nthe aspect of \u201cfair dealing\u201d would have to be considered together and cannot\nbe considered independently.\n&#8211; In the present case, the defence taken by Open AI is under Section\n52(1)(a)(i) of the Copyright Act.\n\n\n## 63\nThe Chancellor, Masters &#038; Scholars of University of Oxford v. Rameshwari Photocopy Services (DB),\n2016 SCC OnLine Del 6229, \u00b6 25.\n\n\n\nCS(COMM) 1028\/2024     Page 92 of 135\n\nB. Legislative History of Section 52(1)(a)\n&#8211; Section 52 forms part of the Copyright Act from its inception in 1957.\nThereafter, it has undergone multiple amendments, with the last amendment\nin 2012.\n&#8211; In the Copyright Act of 1957, Section 52(1)(a) as originally enacted,\nread as under:\n\n\u201c(a) a fair dealing with a literary, dramatic, musical or artistic\nwork for the purposes of \u2013\n(i) research or private study;\n(ii) criticism or review, whether of that work or of any other\nwork.\u201d\n[Emphasis supplied]\n\n&#8211; Section 52(1)(a) was amended in 1994 to read as under:\n\u201c(a) a fair dealing with a literary, dramatic, musical or artistic\nwork, not being a computer programme, for the purposes of \u2013\n(i) private use, including research;\n(ii) criticism or review, whether of that work or of any other\nwork.\u201d\n[Emphasis supplied]\n\n&#8211; In the \u2018Notes on clauses\u2019 in the amendment bill of 1994, the rationale\ngiven for the amendment was that \u201can unduly narrow interpretation of words\n\u201cprivate study\u201d (as existed) may result in harassment to public.\u201d\n&#8211; Once again, Section 52(1)(a) was amended by Copyright (Amendment)\nAct, 2012. The amended provision which exists as on date, reads as under:\n\u201c(a) a  fair  dealing  with  any  work,  not  being  a  computer\nprogramme, for the purposes of \u2013\n(i) private or personal use, including research;\n(ii) criticism or review, whether of that work or of any other\nwork\n\n\n\nCS(COMM) 1028\/2024     Page 93 of 135\n\n(iii) the  reporting  of  current  events  and  current  affairs,\nincluding the reporting of a lecture delivered to public.\nExplanation.\u2014The storing of any work in any electronic medium for\nthe  purposes  mentioned  in  this  clause,  including  the  incidental\nstorage of any computer programme which is not itself an infringing\ncopy  for  the  said  purposes,  shall  not  constitute  infringement  of\ncopyright.\u201d\n[Emphasis supplied]\n\n&#8211; The purpose of various amendments carried out in Section 52(1)(a) has\nbeen  to  widen  the  scope  of  Section  52(1)(a). More  specifically  in  the\nCopyright  (Amendment)  Act,  2012,  an  explanatory  provision  to  Section\n52(1)(a)  of  the  Copyright  Act  was  incorporated,  which  carves  out  an\nexception  in  respect  of  \u2018digital  reproduction  rights\u2019  to  align  it  with  the\namendment  in  Section  14(a)(i),  recognizing  digital  storage  as  part  of\nreproduction.\n\nC. Interpretation of Section 52(1)(a) of the Copyright Act\n&#8211; It has been vehemently contended on behalf of ANI and intervenors\nsupporting ANI  that  since  Section  52  is  in  the  nature  of  an  exception  to\ncopyright infringement the same should be interpreted narrowly. However,\nOpen AI contends that Section 52 of the Copyright Act has to be construed\nbroadly, fulfilling the objective of the Copyright Act as it has been interpreted\nby various Courts.\n&#8211; While interpreting Section 52(1) of the Copyright Act, a Coordinate\nBench of this Court in Narendera Publishing (supra) held that Section 52 of\nthe Copyright Act has to be construed liberally. The relevant paragraph of the\n\n\n\n\nCS(COMM) 1028\/2024     Page 94 of 135\n\nsaid judgment is set out below:\n\n\u201c33. The doctrine of fair use then, legitimizes the reproduction of a\ncopyrightable  work.  Coupled  with  a  limited  copyright  term,  it\nguarantees not only a public pool of ideas and information, but also\na vibrant public domain in expression, from which an individual can\ndraw  as  well  as  replenish.  Fair  use  provisions,  then  must  be\ninterpreted so as to strike a balance between the exclusive rights\ngranted to the copyright holder, and the often competing interest of\nenriching  the  public  domain.  Section  52  therefore  cannot  be\ninterpreted to stifle creativity, and the same time must discourage\nblatant plagiarism. It, therefore, must receive a liberal construction\nin harmony with the objectives of copyright law. Section 52 of the Act\nonly details the broad heads, use under which would not amount to\ninfringement. Resort,  must,  therefore  be  made  to  the  principles\nenunciated by the courts to identify fair use.\u201d\n\n[Emphasis supplied]\n\n\n&#8211; The matter went up in appeal before the Division Bench. However, the\nDivision bench\n## 64\ndid not disturb the aforesaid finding of Single Bench. In fact,\na holistic reading of the judgment of the Division Bench indicates that Section\n52 should not be read as a proviso or an exception to Section 51. Rather,\nSection 52 has been construed liberally by the Division Bench.\n&#8211; At  this  juncture,  it  would  be  appropriate  to  examine  how  similar\nexceptions to copyright infringement are interpreted internationally. In CCH\nCanadian (supra),  the  Supreme Court of  Canada  held  that  fair  dealing\nexception is an integral part of the Copyright Act (of Canada) rather than\nsimply a defence. Therefore, in order to maintain a proper balance between\nthe  rights  of  copyright  owners  and  users,  it  should  not  be  interpreted\n\n## 64\nSyndicate of the Press of the University of Cambridge v. B.D. Bhandari, 2011 SCC OnLine Del 3215.\n\n\n\nCS(COMM) 1028\/2024     Page 95 of 135\n\nrestrictively. The relevant paragraph from the said judgment is set out below:\n\n\u201cBefore reviewing the scope of the fair dealing exception under the\nCopyright Act, it is important to clarify some general considerations\nabout exceptions to copyright infringement. Procedurally, a defendant\nis required to prove that his or her dealing with a work has been fair;\nhowever,  the  fair  dealing  exception  is  perhaps  more  properly\nunderstood as an integral part of the Copyright Act than simply a\ndefence. Any act falling within the fair dealing exception will not be\nan infringement of copyright. The fair dealing exception, like other\nexceptions in the Copyright Act, is a user\u2019s right. In order to maintain\nthe proper balance between the rights of a copyright owner and users\u2019\ninterests, it must not be interpreted restrictively. As Professor Vaver,\nsupra, has explained, at p. 171: \u201cUser rights are not just loopholes.\nBoth owner rights and user rights should therefore be given the fair and\nbalanced reading that befits remedial legislation.\u201d\u201d\n\n[Emphasis supplied]\n\n&#8211; The  legal  principle  that  emerges  from  a  reading  of  the  aforesaid\njudgments is that Section 52 of the Copyright Act strikes a balance between\nthe exclusive rights granted to the owner of the copyright and the competing\npublic  interest  of  encouraging  creativity  and  disseminating  knowledge.\nHence, Section 52 is not in the nature of a proviso or an exception to the\nSection 51 of the Copyright Act.\n&#8211; In  my  respectful  view,  Section  52  independently  defines  the\nrights\/privileges of a user in respect of copyrighted works and must not be\ninterpreted in a restrictive manner. It is an integral part of the Copyright Act\nand therefore, it has to be given broad and liberal interpretation.\n\nD. Scope of Section 52(1)(a)\n&#8211; To  determine  whether  the  storage  by  Open AI  falls  under  Section\n\n\n\nCS(COMM) 1028\/2024     Page 96 of 135\n\n52(1)(a) of the Copyright Act, a two-step examination is required:\ni. Whether the storage by Open AI is for one of the specific purposes\nmentioned in Section 52(1)(a) of the Copyright Act. (Purpose Test)\nii. Whether  the  storage  can  be  considered  as  \u201cfair  dealing\u201d  or  not.\n(Fairness Test)\n\n## I. PURPOSE TEST\n&#8211; In the present case, it is an admitted position that Open AI electronically\nstores ANI\u2019s original literary works, at least on a temporary basis. Even if\nliterary works of ANI collected and stored by Open AI are deleted after the\ncompletion of training, it would not alter the position that temporary storage\noccurred in the first place.\n&#8211; Hence, to satisfy the first requirement under Section 52(1)(a) of the\nCopyright Act, it has to be determined whether the purpose for which Open\nAI  has  stored  ANI\u2019s  literary  works  falls  under  Section  52(1)(a)  of  the\nCopyright Act. The  relevant  extracts  from  Section  52(1)(a)  providing  the\npurposes mentioned therein are set out below:\n\u201c52.  Certain  acts  not  to  be  infringement  of  copyright.\u2014(1)  The\nfollowing  acts  shall  not  constitute  an  infringement  of  copyright,\nnamely\u2014\n(a) a fair dealing with any work, not being a computer programme, for\nthe purposes of \u2013\n(i) private or personal use, including research;\n(ii) criticism or review, whether of that work or of any other\nwork;\n(iii) the  reporting  of  current  events  and  current  affairs,\nincluding the reporting of a lecture delivered to public.\nExplanation.\u2014The storing of any work in any electronic medium for the\npurposes mentioned in this clause, including the incidental storage of\n\n\n\nCS(COMM) 1028\/2024     Page 97 of 135\n\nany computer programme which is not itself an infringing copy for the\nsaid purposes, shall not constitute infringement of copyright.\u201d\n[Emphasis supplied]\n&#8211; On a plain reading of Section 52(1)(a), it is manifest that sub-clauses\n(i), (ii) and (iii) of Section 52(1)(a) are disjunctive in nature. Therefore, if the\npurpose falls under any of the three sub-clauses of Section 52(1)(a), the use\nwould be entitled to protection against an infringement action. In the present\ncase, Open AI claims protection under Section 52(1)(a)(i) of the Copyright\nAct, i.e. \u201cprivate or personal use, including research\u201d.\n&#8211; In  support  of  its  contention  that  Open  AI  does  not  fulfil  the\nrequirements of the purpose test under Section 52(1)(a)(i), ANI has made the\nfollowing submissions:\na. The private use\/research cannot be for commercial purposes.\nb. The first copy of the work has to be a non-infringing copy.\nc. The use by Open AI does not amount to \u201cprivate or personal use,\nincluding research\u201d\nI will deal with each of these separately.\na. Whether  commercial  use  is  entitled  to  protection  under  Section\n## 52(1)(a)(i)\n&#8211; On behalf of ANI, it is contended that the phrase \u201cprivate or personal\nuse, including research\u201d under Section 52(1)(a)(i) of the Copyright Act would\nnot be applicable in the present case since ANI\u2019s works are being used by\nOpen AI for commercial purposes and to earn profits. Moreover, Open AI is\na business venture which profits from large scale reproduction, adaptation and\ndistribution of copyrighted content.\n\n\n\nCS(COMM) 1028\/2024     Page 98 of 135\n\n&#8211; Per Contra, it is contended by Open AI that limiting the expression\n\u201cprivate\u201d use including \u201cresearch\u201d only for non-commercial purposes and to\nindividuals would amount to a narrow interpretation of Section 52 (1)(a) of\nthe Copyright Act. According to Open AI, wherever non-commercial use is\nintended, it has been specifically provided in the statute.\n&#8211; It has been argued on behalf of the intervenors IMI and DNPA that since\n## Section  52(1)(ad)\n## 65\nof  the  Copyright  Act  makes  a  reference  to  \u201cnon-\ncommercial personal use\u201d, the said limitation should also be read in Section\n52(1)(a) of the Act.\n&#8211; I am unable to agree with the said submission. A perusal of various\nprovisions of Section 52(1) of the Copyright Act clearly demonstrate that the\nlegislature  was  aware  of  the  distinction  between  \u2018commercial\u2019  and  \u2018non-\ncommercial\u2019 use and thought it fit to exclude commercial use only in specific\ncases.\n&#8211; Wherever the protection under Section 52(1) of the Copyright Act is\nprovided only for non-commercial use, it has been specifically provided. For\ninstance,  Section  52(1)(ad)  makes  it  clear  that  the  protection  therein  is\navailable  only  for  \u201cnon-commercial  personal  use\u201d.  Similarly,  Section\n\n## 65\n(ad) the making of copies or adaptation of the computer programme from a personally legally obtained\ncopy for non-commercial personal use;\n\n\n\n\nCS(COMM) 1028\/2024     Page 99 of 135\n\n## 52(1)(k)(ii)\n## 66\n## , Section 52(1)(l)\n## 67\n## , Section 52(1)(n)\n## 68\nand Section 52(1)(o)\n## 69\nof\nthe Copyright Act, specifically, stipulate that the benefit of the exceptions\ncontained therein would be available for non-commercial purposes only. The\nrequirement  of  non-commercial  use  is  conspicuously  absent  in  Section\n## 52(1)(a).\n&#8211; The  defence  under  Section  52(1)(a)  of  the  Copyright  Act  is  not\nrestricted to non-commercial use can also be gauged from the fact that a\nperson may be writing a book review that could be done on a commercial\nbasis, and yet be covered under Section 52(1)(a)(ii) of Copyright Act. In\nSuper Cassettes v. Hamar Televisions (supra), while dealing with the defence\nunder Section 52(1)(a)(ii) of the Copyright Act, it was held that commercial\nuse of a copyrighted work would not make it unfair or take it away from the\nambit of fair dealing.\n&#8211; Similarly, even \u201creporting of current events and current affairs\u201d can be\nfor  commercial  purposes.  For  example,  a  journalist  or  a  news  agency\nreporting a current event, may be a paid activity and yet the journalist or the\nnews agency would be entitled to take the defence under Section 52(1)(a)(iii)\nof the Copyright Act. Therefore, in my view, the limitation of non-commercial\nuse  cannot  be  read  into  Section  52(1)(a)(i).  Merely  because  the  use  of  a\ndefendant is commercial would not mean that the defendant would not be\n\n## 66\n(k) the causing of a recording to be heard in public by utilising it, \u2014\n## (i) &#8230;\n(ii) as part of the activities of a club or similar organisation which is not established or conducted for profit;\n## 67\n(l) the performance of a literary, dramatic or musical work by an amateur club or society, if the performance\nis given to a non-paying audience, or for the benefit of a religious institution;\n## 68\n(n) the storing of a work in any medium by electronic means by a non-commercial public library, for\npreservation if the library already possesses a non-digital copy of the work;\n## 69\n(o) the making of not more than three copies of a book (including a pamphlet, sheet of music, map, chart\nor plan) by or under the direction of the person in charge of a non-commercial public library for the use of\nthe library if such book is not available for sale in India;\n\n\n\nCS(COMM) 1028\/2024     Page 100 of 135\n\nentitled to take the defences provided under Section 52 of the Copyright Act\nunless there is a specific bar on commercial use as stipulated in some of the\nprovisions of Section 52 of the Copyright Act.\n&#8211; In this regard, a reference may be made to the judgment of the Canadian\nSupreme Court in CCH Canadian (supra). In the said case, the Canadian\nSupreme Court was dealing with Section 29\n## 70\nof Copyright Act of Canada\n(R.S.C., 1985, c. C-42) which is pari materia with Section 52(1)(a) of the\nCopyright  Act.  It  was  held  that  \u201cresearch\u201d  should  be  given  a  liberal\ninterpretation and that research should not be confined to non-commercial\nuse. The relevant paragraph of the said judgment is set out below:\n\n\u201c51. The fair dealing exception under s. 29 is open to those who can\nshow that their dealings with a copyrighted work were for the purpose\nof research or private study. \u201cResearch\u201d must be given a large and\nliberal interpretation in order to ensure that users\u2019 rights are not\nunduly constrained. I agree with the Court of Appeal that research is\nnot  limited  to  non-commercial  or  private  contexts.  The Court of\nAppeal correctly noted, at para. 128, that \u201c[r]esearch for the purpose\nof advising clients, giving opinions, arguing cases, preparing briefs\nand  factums  is  nonetheless  research.\u201d  Lawyers  carrying  on  the\nbusiness of law for profit are conducting research within the meaning\nof s. 29 of the Copyright Act.\u201d\n[Emphasis supplied]\n&#8211; Other common law jurisdictions have also taken a similar view that\nmerely because the use is a commercial one, may not take it outside the\ndefence of \u201cfair dealing\u201d\n## 71\n## .\n\n## 70\nResearch, private study, etc.\n&#8211; Fair dealing for the purpose of research, private study, education, parody or satire does not infringe\ncopyright.\n## 71\nSociety of Composers, Authors and Music Publishers of Canada v. Bell Canada, 2012 SCC 36; Television\nNew Zealand Ltd v. Newsmonitor Services Ltd, [1994] 2 NZLR 91.\n\n\n\nCS(COMM) 1028\/2024     Page 101 of 135\n\n&#8211; ANI has relied upon paragraph 65 of B.D. Bhandari (supra) to argue\nthat the defences under Section 52 of the Copyright Act would not be available\nfor commercial use. The relevant paragraph of B.D. Bhandari (supra) relied\nupon by ANI is set out below:\n\n\u201c65. It is clear from the reading of the aforesaid provision that this\nwould not come to the aid of the respondents who have published the\nguide for commercial exploitation. In view of the law laid down in\nvarious cases, commercial exploitation will take the matter outside the\nambit of \u2018fair use\u2019. The defence of \u2018fair use\u2019 as provided under Section\n52(1)(h)  of  the Act  is  only  available  to  the  teacher  and  pupil  to\nreproduce  the  literary  work  in  the  course  of  instructions  or\nexamination paper setter to reproduce the literary work as part of the\nquestions or to the pupil to reproduce the literary work as answers to\nsuch questions. The respondents are neither teachers nor students nor\na person giving or receiving instruction. The defence of \u2018fair use\u2019\nunder Section 52(1)(h) of the Act only provided to a teacher and pupil\nand not otherwise. If this defence is allowed to a publisher, then it\nwould result in a situation where every publisher, without permission\nfrom the owner of copyright, would reproduce the ad verbatim literary\ntext from the educational textbooks prescribed by the University into\ntheir books and exploit the same for commercial gains and benefits\nand that too without giving any royalty or payment to the right owners\ntowards such reproduction. This practice would obviously discourage\ncreativity of authors who put their skill, labour, years of knowledge,\nexpertise, etc. into these educational books and encourage infringers\nlike  the  respondents  to  reproduce  ad  verbatim  literary  text  from\noriginal publications into their infringing books on the pretext of \u2018fair\nuse\u2019. Obviously,  therefore,  Section  52(1)(h)  of  the  Act  has  no\napplication to the facts of this case.\u201d\n[Emphasis supplied]\n\n&#8211; The aforesaid observations made in B.D. Bhandari (supra) were in the\ncontext of Section 52(1)(h) of the Copyright Act [now Section 52(1)(i) of the\nCopyright Act]. For ease of reference, Section 52(1)(i) of the Copyright Act\n\n\n\nCS(COMM) 1028\/2024     Page 102 of 135\n\nis set out below:\n\u201c(i) the reproduction of any work\u2014\n(i) by a teacher or a pupil in the course of instruction; or\n(ii) as part of the questions to be answered in an examination; or\n(iii) in answers to such questions;\u201d\n\nThe scope of Section 52(1)(i) is entirely different from Section 52(1)(a) as\nSection  52(1)(i)  deals  with  reproduction  of  any  work  in  the  \u201ccourse  of\ninstruction\u201d or \u201cas part of the questions to be answered in an examination\u201d or\n\u201cin answers to such questions\u201d. The Division Bench held that the defence\nunder Section 52(1)(h) is only applicable to a teacher and pupil and not to a\npublisher.\n&#8211; Ultimately, the Division Bench held that the act of the defendants was\nnon-infringing and amounted to creating a \u2018derivative work\u2019 of distinctive\ncharacter.  In B.D. Bhandari (supra),  the  guidebooks  were  being  sold  by\ndefendants for commercial purposes and yet the act of the defendants was held\nto be non-infringing.\n&#8211; In Rameshwari Photocopy Services (DB) (supra), while dealing with\nthe defence under Section 52(1)(h) [now section 52(1)(i) of the Copyright\nAct], the Division Bench gave the benefit of Section 52(1)(h) to the defendant,\neven though the use of the defendant was for commercial purposes. It was\nheld that a teacher or a pupil cannot be expected to buy photocopiers and the\nargument that there cannot be an intermediary for photocopying, was rejected.\n&#8211; The plaintiff has also placed reliance on paragraph 21 of the judgment\nof this Court in Jiwan Publishing House (supra), in support of its contention\nthat the defence under Section 52(1)(a) of the Copyright Act would not be\navailable to a publisher who publishes a book for commercial exploitation.\n\n\n\nCS(COMM) 1028\/2024     Page 103 of 135\n\nThe relevant paragraph of the said judgment is set out below:\n\u201c21. In the written statement of defendants 1 and 2 vide para 4 of the\npreliminary objections, as also in the reply to the application, reference\nhas  been  made  to  Section  52(1)(a)(i)  and  S  52(1)(b)(iii)  of  the\nCopyright Act as providing the defences available to defendants 1 and\n&#8211; In the Copyright Act, there is nothing like sub clause (iii) to be found\nunder  clause  (b)  of  sub-section  (1)  of  Section  52  of  the Act.  That\nreference appears to be mistaken. It would, therefore, suffice to deal\nwith the plea raised by reference to S. 52(1)(a)(i) of the Act, which\nprovides inter alia, that a fair dealing with the literary work for the\npurpose of research or private study does not constitute infringement of\ncopyright. The words \u2018research or private study\u2019 have been substituted\nby  the  words  \u2018private  use  including  research\u2019  by  the  Copyright\n(Amendment) Act, 1994 (Act 381 of 1994). What is contemplated is a\ndefence to the person conducting research or private study who while\ndoing so, if dealing fairly with a literary work, may not incur wrath of\nthe copyright having been infringed. But, if a publisher publishes a\nbook  for  commercial  exploitation  and  in  doing  so  infringes  a\nCopyright,  the  defence  under  section  52(1)(a)(i)  would  not  be\navailable to such a publisher though the book published by him may\nbe used or be meant for use in research or private study. The defence\nraised by defendants 1 and 2 based on Section 52(1) (a) (i) is not\navailable to them and the plea so raised has to be rejected.\u201d\n\n[Emphasis supplied]\n\n&#8211; In Jiwan Publishing House (supra), the defendants were publishing\nprevious years\u2019 question papers of the examinations conducted by Central\nBoard of Secondary Education (\u2018CBSE\u2019), without taking any license from\nCBSE. The plaintiff contended that it was the exclusive licensee of CBSE in\nrespect of reprinting and reproduction of the question papers and actions of\nthe defendants were in breach of plaintiff\u2019s rights. In those circumstances, it\nwas held that the defence under Section 52(1)(a)(i) of the Copyright Act is\navailable  to  a  person  conducting  \u201cresearch\u201d  or  \u201cprivate study\u201d. The  said\ndefence would not be available to a publisher whose book may be used for\n\n\n\nCS(COMM) 1028\/2024     Page 104 of 135\n\n\u201cresearch\u201d or \u201cprivate study\u201d by third parties such as students.\n&#8211; ANI has also placed reliance on the judgment of the Bombay High\nCourt in TIPS Industries v. Wynk Music (supra) in support of its contention\nthat where the defendants commercially exploit copyrighted material of the\nplaintiff, it would not be entitled to the defence under Section 52(1)(a) of the\nCopyright Act. Once again, in this case, there was no \u201cprivate use\u201d of the\ncopyrighted material by the defendants. On the contrary, the defendants were\nmaking available the plaintiff\u2019s music to third party users on a commercial\nbasis through online streaming services and the defence taken was that the\nsaid users were using the content in a private or personal capacity.\n&#8211; In the present case, as is explained hereinafter\n## 72\n, the use by Open AI is\nonly for the purposes of training its LLM, which is completely an internal\nprocess and does not involve providing the copyrighted material of ANI to\nany third party. Even after the training is complete, the training data is never\nmade available either in its natural language or tokenized form to any person.\nTherefore,  the  judgments  in Jiwan  Publishing House (supra)  and TIPS\nIndustries v. Wynk (supra) are of no assistance to ANI in the present case.\n&#8211; In light of the discussion above, in my view, merely because the use of\nOpen AI is for a commercial purpose would not mean that Open AI is not\nentitled to take the defence under Section 52(1)(a) of the Copyright Act.\n\nb. Requirement of \u201cnon-infringing copy\u201d under Section 52(1)(a) of the\n## Copyright Act.\n&#8211; The Explanation to Section 52(1)(a) of the Copyright Act makes it clear\n\n## 72\nRefer to sub-heading \u2018Whether use by Open AI amounts to \u201cprivate or personal use, including research\u201d\u2019.\n\n\n\nCS(COMM) 1028\/2024     Page 105 of 135\n\nthat  storage  (temporary\/permanent)  of  a  work  in  electronic  means  for\npurposes  mentioned  in  Section  52(1)(a)  of  the  Copyright  Act  shall  not\nconstitute infringement of copyright subject to the test of \u201cfair dealing\u201d. The\nExplanation to Section 52(1)(a) of the Copyright Act has been the subject\nmatter of intense debate between the counsel for the parties. For the sake of\nconvenience, Section 52(1)(a) is set out below:\n\n\u201c52.  Certain  acts  not  to  be  infringement  of  copyright.\u2014(1)  The\nfollowing  acts  shall  not  constitute  an  infringement  of  copyright,\nnamely\u2014\n(a) a fair dealing with any work, not being a computer programme, for\nthe purposes of \u2013\n(i) private or personal use, including research;\n(ii) criticism or review, whether of that work or of any other\nwork\u2019\n(iii) the  reporting  of  current  events  and  current  affairs,\nincluding the reporting of a lecture delivered to public.\nExplanation.\u2014The storing of any work in any electronic medium for\nthe purposes mentioned in this clause, including the incidental storage\nof any computer programme which is not itself an infringing copy for\nthe said purposes, shall not constitute infringement of copyright.\u201d\n[Emphasis supplied]\n\n&#8211; The  submission  made  on  behalf  of  Open AI  is  that  the  use  of  \u2018,\u2019\n(commas) before and after the expression \u201cincluding the incidental storage of\nany computer programme which is not itself an infringing copy for the said\npurposes\u201d makes it clear that the limitation of not being an infringing copy is\napplicable only in respect of a computer programme.\n&#8211; It is the submission of ANI as well as the intervenors, DNPA and IMI\nthat Open AI stores a copy of ANI\u2019s works that has been obtained in an\nunauthorised manner. The storage of an infringing copy of a work would not\n\n\n\nCS(COMM) 1028\/2024     Page 106 of 135\n\nqualify  as  a  defence  in  light  of  Explanation  to  Section  52(1)(a)  of  the\nCopyright Act. Any other interpretation of the Explanation would expand the\nscope of Section 52(1)(a) of the Copyright Act as \u201ccomputer programmes\u201d\nhave already been excluded from the main text of Section 52(1)(a).\n&#8211; No doubt \u201ccomputer programmes\u201d have been excluded from the main\ntext of Section 52(1)(a) of the Copyright Act. What the Explanation refers to\nis  the  \u201cincidental  storage  of  any  computer  programme\u201d  i.e.  computer\nprogramme is incidental to the electronic storage of a work. The limitation of\nnot being an infringing copy is, therefore, only applicable to the incidental\nstorage  of  computer  programmes.  This  position  is  quite  clear  from  the\npositioning of the \u2018,\u2019 (commas) in the Explanation.\n&#8211; The  Explanation  to  Section  52(1)(a)  does  not  expand  the  scope  of\nprotection with respect to \u201ccomputer programmes\u201d and merely clarifies that\nstorage  of  an  infringing  copy  of  a  computer  programme  would  not  be\nprotected even if such storage is incidental. Therefore, ANI\u2019s submission with\nrespect to expansion of scope of Section 52(1)(a) cannot be accepted.\n&#8211; In  Sections  52(1)(aa),  52(1)(ab)  and  52(1)(ad),  the  words  \u201clawful\u201d\/\n\u201clegally  obtained  copy\u201d  has  been  used  only  in  respect  of  computer\nprogrammes. Therefore, it cannot be said that \u201cnon-infringing copy\u201d would\napply to storage of other works unless it has been specifically mentioned. For\ninstance, in Section 52(1)(c)\n## 73\nit has been specifically mentioned that for the\n\n## 73\n(c) transient or incidental storage of a work or performance for the purpose of providing electronic links,\naccess or integration, where such links, access or integration has not been expressly prohibited by the right\nholder, unless the person responsible is aware or has reasonable grounds for believing that such storage\nis of an infringing copy:\nProvided that if the person responsible for the storage of the copy has received a written complaint from the\nowner of copyright in the work, complaining that such transient or incidental storage is an infringement, such\nperson responsible for the storage shall refrain from facilitating such access for a period of twenty-one days\nor till he receives an order from the competent court refraining from facilitating access and in case no such\n\n\n\nCS(COMM) 1028\/2024     Page 107 of 135\n\npurposes  of  providing  electronic  links,  the  person  responsible  must  have\nreasonable grounds to believe that an infringing copy is not being stored.\n&#8211; If the interpretation given by ANI is accepted, it would imply that a\nresearcher going to a public library and photocopying a page from a book for\nthe purposes mentioned under Section 52(1)(a) of the Copyright Act would\nbe entitled to defence of fair dealing. However, if the same researcher scans\nthe said page electronically and stores the same in its device for the same\npurpose, the protection would not be available.\n&#8211; In my considered view, this distinction between storage in electronic\nmedium and storage in the physical medium cannot be the intent of Section\n52(1)(a)  of  the  Copyright Act. Therefore,  I  do  not  find  any  merit  in  the\ncontention that since Open AI stores an infringing copy of ANI\u2019s work, it\nwould not be entitled to defence under Section 52(1)(a).\n&#8211; In Bartz v. Anthropic (supra), the US District Court clearly made a\ndistinction  between  acquiring  plaintiff\u2019s  books  in  a  lawful  manner  and\nobtaining copies from \u2018Shadow Libraries\u2019. The US Court held that scanning\nof lawfully acquired books amounted to fair use, whereas obtaining copies\nfrom shadow libraries was not fair use.\n&#8211; In the present case, it is not the case of ANI that Open AI obtains copies\nof ANI\u2019s works from unauthorized sources or by breaking through the paywall\nof ANI. It is not disputed that Open AI obtains copies from information\/data\nthat is freely available on ANI\u2019s website. Therefore, in any event, it cannot be\nsaid that Open AI has obtained an infringing copy of ANI\u2019s works.\n&#8211; Therefore, in my view, under the Explanation to Section 52(1)(a) the\n\norder is received before the expiry of such period of twenty-one days, he may continue to provide the facility\nof such access;\n\n\n\nCS(COMM) 1028\/2024     Page 108 of 135\n\nlimitation of not being an infringing copy is applicable only in respect of a\n\u2018computer programme\u2019 and not in respect of the works stored in electronic\nmode.\n\nc. Whether  use  by  Open AI  amounts  to  \u201cprivate or personal  use,\nincluding research\u201d\n&#8211; To begin with, as noted in my findings in Issue no. 2, the cut-off dates\nfor storage and training of the LLM models underlying ChatGPT were April\n2022  (for  GPT  4)  and  April  2024  (for  GPT  4o),  which  is  prior  to  the\npublication of articles used by ANI as instances of reproduction. During the\ncourse of hearing, ANI has not provided any further instances to show the\nextracted data from ANI\u2019s website has been used for training. Therefore, there\nis no factual foundation in the plaint to reach a finding of infringement in\nrespect of the training claim. However, since it is an admitted position that the\noriginal literary works of ANI are stored by Open AI at least temporarily\nduring the course of training\n## 74\n, this Court shall examine whether the storage\nof literary works by Open AI for training LLMs underlying ChatGPT qualifies\nas \u201cprivate or personal use, including research\u201d under Section 52(1)(a) of the\n## Copyright Act.\n&#8211; Open AI  submits  that  the  purpose  of  the  storage  of ANI\u2019s  original\nliterary works falls within of Section 52(1)(a)(i) i.e. \u201cprivate or personal use,\nincluding research\u201d. More particularly, Open AI\u2019s electronic storage of ANI\u2019s\noriginal literary works for training LLMs underlying ChatGPT, would fall\n\n## 74\nParagraph 11.3 of Response on behalf of the Defendant to the written submissions on behalf of Prof. Arul\nGeorge Scaria, Amicus Curiae; Paragraphs 7, 13.1, 58.3 of Response on behalf of the Defendant to the written\nsubmissions  on  behalf  of  Mr.  Adarsh  Ramanujan,  Amicus  Curiae;  Paragraphs  63,  76  of  the  Written\nSubmissions on behalf of the Defendant dated 28\nth\n## March, 2025.\n\n\n\nCS(COMM) 1028\/2024     Page 109 of 135\n\nwithin the ambit of \u201cprivate\u201d use, including \u201cresearch\u201d.\n&#8211; It is the submission of ANI duly supported by Mr. Adarsh Ramanujan\nthat \u201cprivate use\u201d under Section 52(1)(a)(i) of the Copyright Act is applicable\nonly to an individual\/human being.\n&#8211; The  interpretation  advanced  by  Mr.  Ramanujan  is  that  the  words\n\u201cprivate\u201d and \u201cpersonal\u201d are overlapping and they draw meaning from each\nother. Therefore, the interpretative rule of noscitur a sociis would require that\nthe meaning of \u201cprivate\u201d and \u201cpersonal\u201d be interpreted in an analogous and\ncognate sense.\n&#8211; According to Prof. Scaria, the Copyright Act does not define the term\n\u201cprivate\u201d. He refers to the definition of \u201cprivate\u201d given in Meriam Webster\nDictionary and Cambridge Dictionary to submit that the term \u201cprivate\u201d cannot\nbe confined to individual persons.\n&#8211; The Merriam Webster Dictionary gives the meaning of \u201cprivate\u201d as\n&#8211; \u201cintended for or restricted to the use of a particular person, group, or class\u201d\nand \u201cbelonging to or concerning an individual person, company, or interest\u201d.\nThe Cambridge Dictionary defines \u201cprivate\u201d as \u201conly for one person or\ngroup and not for everyone\u201d.\n&#8211; From  the  aforesaid  definitions,  it  transpires  that  the  term  \u201cprivate\u201d\ncannot be confined to an individual and can apply to a closed group or a\ncompany. The expression \u201cpersonal\u201d may be confined to individual persons,\nhowever, the term \u201cprivate\u201d would include other private entities, including\nprivate companies.\n&#8211; As  can  be  discerned  from  the  amendments  carried  out  in  Section\n52(1)(a) from time-to-time, the legislature has throughout been expanding the\nscope of Section 52(1)(a) of the Copyright Act. The latest amendment of 2012\n\n\n\nCS(COMM) 1028\/2024     Page 110 of 135\n\nprovides  that  both  \u201cprivate\u201d  as  well  as  \u201cpersonal\u201d  use  are  eligible  for\nprotection under Section 52(1)(a) of the Copyright Act. If the submission of\nMr. Ramanujan that \u201cprivate\u201d and \u201cpersonal\u201d are to be used interchangeably,\nis accepted, then the word \u201cprivate\u201d would be rendered otiose.\n&#8211; The view that the term \u201cprivate\u201d should be given an expansive meaning\nalso finds support from the judgment of the Supreme Court in B. Malini\nMallya (supra). The respondent-plaintiff therein claimed copyright ownership\nover  a  ballet  dance  form  which  was  performed  by  the  performers  at  the\nappellant\/defendant\u2019s institute in New Delhi without any permission. In this\ncase, the  Supreme Court gave the benefit of  Section 52  to the defendant\ninstitute by holding that a dance performance at an educational institution\nwould  be  covered  under  Section  52(1)(a)(i)  of  the  Copyright  Act  as  it\namounted  to  \u201cprivate use\u201d. The  relevant  extracts  from B. Malini Mallya\n(supra) are set out below:\n\n\u201c38. The High Court, in our opinion, should have clarified that the\nappellants  can  also  take  the  statutory  benefit  of  the  provisions\ncontained in clauses (a), (i) and (l) of sub-section (1) of Section 52 of\nthe Act.\n&#8211; Section 52 of the Act provides for certain acts which would not\nconstitute an infringement of copyright. When a fair dealing is made,\ninter alia, of a literary or dramatic work for the purpose of private use\nincluding research and criticism or review, whether of that work or of\nany other work, the right in terms of the provisions of the said Act\ncannot be claimed. Thus, if some performance or dance is carried out\nwithin the purview of the said clause, the order of injunction shall not\nbe applicable.\n&#8211; Similarly, the appellant being an educational institution, if the\ndance is performed within the meaning of provisions of clause (i) of\nsub-section (1) of Section 52 of the Act strictly, the order of injunction\nshall  not  apply  thereto  also.  Yet  again,  if such  performance  is\nconducted before a non-paying audience by the appellant, which is an\n\n\n\nCS(COMM) 1028\/2024     Page 111 of 135\n\ninstitution if it comes within the purview of amateur club or society,\nthe  same  would  not  constitute  any  violation  of the  said  order  of\ninjunction.\u201d\n[Emphasis supplied]\n\n&#8211; In the present case, Open AI stores the literary works in a closed space\nwithout access to the public. The data obtained by the LLMs for training\npurposes is used for private purposes. The said data is accessible only to the\nLLM models themselves. The said data is not publicly available to any human\nentity either for access or for download. Therefore, in my opinion, the use\namounts to being purely private.\n&#8211; The term \u201cresearch\u201d is not defined under the Copyright Act. However,\nMerriam-Webster Dictionary gives the meaning of \u201cresearch\u201d as \u201cto study or\ninvestigate carefully\u201d.\n&#8211; The Madras High Court in Blackwood v. A.N. Parasuraman (supra),\nused the dictionary meaning of \u201cresearch\u201d for the purpose of interpreting\nSection 52(1)(a) of the Copyright Act. The relevant paragraphs of the said\njudgment are reproduced here below:\n\n\u201c154. There is no definition of the word \u201cresearch\u201d in the Act, and,\ntherefore its meaning has to be ascertained from ordinary English,\nusage. In Funk and Wagnalls English dictionary \u201cresearch\u201d is said to\nmean, I am extracting that meaning which appears to be most relevant\nto the present context, \u201cDiligent protracted investigation especially for\nthe purpose of adding to human knowledge, studious enquiry.\u201d\n&#8211; In \u201cThe Shorter Oxford Dictionary\u201d the meaning attributed is \u201cAn\ninvestigation directed to the discovery of some fact by careful study of\na subject; investigation, inquiry into things.\u201d&#8230;\u201d\n[Emphasis supplied]\n\n&#8211; The process of \u201cresearch\u201d is generally an intermediary process in all\ncases. It is undertaken before an output is generated. Research happens before\n\n\n\nCS(COMM) 1028\/2024     Page 112 of 135\n\nan academician writes a book or a research paper, or before a judge writes a\njudgment, or before a medicinal chemist develops a drug, or before any such\ndevelopmental process. It is normally a closed activity and is not disclosed to\nthe general public. It is always the output of the research that is communicated\nto the public.\n## 75\n\n&#8211; From the submissions made in the present case, it transpires that the\nprocess of training LLMs underlying ChatGPT involves machine learning of\nthe stored literary works by screening and organising them. In this process,\nthe  data  is  analysed  by  making  extractions  from  the  literary  works  and\nconverting them into machine-readable training inputs. This data is used to\ntrain the LLMs iteratively to rectify errors and generate the most probable\nstatistical output. Accordingly, the process of training of LLMs underlying\nChatGPT  can  be  considered  as  a  form  of  research  for  generating  new\nknowledge and advancing artificial intelligence systems.\n&#8211; Clearly,  when  Section  52(1)(a)  was  last  amended  in  2012,  the\nlegislature could not have imagined the advent of artificial intelligence (AI)\nand\/or LLMs. However, when the Court interprets the aforesaid provisions in\nthe light of modern-day technological developments, the Court has to give a\nliberal  and  purposive  interpretation  using  the  \u2018doctrine  of  updating\nconstruction\u2019.\n&#8211;  The doctrine of updating construction gives effect to the true intention\nof  the  legislature  by  interpreting  words  in  a  statute  in  accordance  with\nchanging social conditions and developments in science and technology. This\ndoctrine was applied by a Constitutional Bench of Supreme Court in State\n\n## 75\nMerriam  Webster  Dictionary  defines  the  verb  \u201cresearch\u201d  as  \u201cto  study  or  investigate  carefully\u201d\n## <https:\/\/www.merriam-webster.com\/dictionary\/research.>.\n\n\n\nCS(COMM) 1028\/2024     Page 113 of 135\n\n(Through  CBI\/  New  Delhi) v.  S.J.  Choudhary\n## 76\n.  In  the  said  case,  the\nSupreme Court was dealing with a question of law i.e. \u201cwhether the opinion\nof a typewriter expert is admissible in evidence under Section 45 of the Indian\nEvidence Act, 1872\u201d. The Supreme Court allowed the appeal and observed\nthat  at  the  time  of  enactment  of  the  Indian  Evidence  Act  in  1872,  the\ntypewriters  were  \u201cpractically  unknown\u201d.  Hence,  typewriting  was  not\nspecifically mentioned in Section 45 of Indian Evidence Act. The Supreme\nCourt relied upon Francis Bennion\u2019s Statutory Interpretation, 2nd  Edn., to\ngive a wider meaning to the expression \u201cscience\u201d used in Section 45\n## 77\nof the\nIndian Evidence Act, to hold that the opinion of an expert in typewriting is\nadmissible. The relevant paragraph of S.J. Choudhary (supra) which quotes\nBennion\u2019s Statutory Interpretation is reproduced below:\n\n\u201c10.  Statutory  Interpretation by  Francis  Bennion,  2nd  Edn.,  Section\n288 with the heading \u201cPresumption that updating construction to be\ngiven\u201d states one of the rules thus: (p. 617)\n## \u201c***\n(2) It is presumed that Parliament intends the court to apply to an\nongoing Act a construction that continuously updates its wording\nto  allow  for  changes  since  the  Act  was  initially  framed  (an\nupdating construction). While it remains law, it is to be treated as\nalways  speaking.  This  means  that  in  its  application  on  any  date,\nthe language of the Act, though necessarily embedded in its own\ntime, is nevertheless to be construed in accordance with the need\nto treat it as current law.\n## ***\u201d\n\n## 76\n## (1996) 2 SCC 428.\n## 77\n&#8211; Opinions of experts. &#8211; When the Court has to form an opinion upon a point of foreign law or of science,\nor art, or as to identity of handwriting, or finger impressions, the opinions upon that point of persons specially\nskilled in such foreign law, science or art, or in questions as to identity of handwriting or finger impressions\nare relevant facts.\nSuch persons are called experts.\n\n\n\nCS(COMM) 1028\/2024     Page 114 of 135\n\nIn the comments that follow it is pointed out that an ongoing Act is taken\nto be always speaking. It is also, further, stated thus: (pp. 618-19)\n\u201cIn construing an ongoing Act, the interpreter is to presume that\nParliament  intended  the  Act  to  be  applied  at  any  future  time  in\nsuch  a  way  as  to  give  effect  to  the  true  original  intention.\nAccordingly the interpreter is to make allowances for any relevant\nchanges that have occurred, since the Act\u2019s passing, in law, social\nconditions, technology, the meaning of words, and other matters.\nJust as the US Constitution is regarded as \u2018a living Constitution\u2019, so\nan ongoing British Act is regarded as \u2018a living Act\u2019. That  today\u2019s\nconstruction involves the supposition that Parliament was catering\nlong ago for a state of affairs that did not then exist is no argument\nagainst  that  construction.  Parliament,  in  the  wording  of  an\nenactment,  is  expected  to  anticipate  temporal  developments.  The\ndrafter will try to foresee the future, and allow for it in the wording.\n## ***\nAn enactment of former days is thus to be read today, in the light\nof   dynamic   processing   received   over   the   years,   with   such\nmodification  of  the  current  meaning  of  its  language  as  will  now\ngive  effect  to  the  original  legislative  intention.  The  reality  and\neffect of dynamic processing provides the gradual adjustment. It is\nconstituted  by  judicial  interpretation,  year  in  and  year  out.  It  also\ncomprises processing by executive officials.\u201d \u201d\n[Emphasis supplied]\n\n&#8211; In my opinion, the expression \u201cresearch\u201d should be given an updating\nconstruction by taking into account the modern-day technical advancements.\nWith the advent of these technologies, research\/ learning is no longer confined\nto humans. It is now being done through Artificial Intelligence. However,\nultimately  the  research  is  at  the  behest  of  humans  and  for  the  benefit  of\nhumans. For example, Section 52(1)(i) of the Copyright Act exempts the act\nof reproduction by a teacher \u201cin the course of instruction\u201d from the ambit of\n\n\n\nCS(COMM) 1028\/2024     Page 115 of 135\n\ninfringement. If tomorrow a human teacher is replaced by an AI bot or Robot\n## 78\n\nto say that the said exception could only be used by a human would be a\nregressive view. Such an approach would limit societal progress. Therefore,\nthe acts of further research cannot be confined to acts of human being alone\nand the same would extend to machine learning as well.\n&#8211; Therefore, on a prima facie view, from the above analysis, this Court is\nof the view that the process of training LLMs underlying ChatGPT undertaken\nby Open AI using stored literary work of ANI falls under \u201cprivate or personal\nuse, including research\u201d as provided in Section 52(1)(a) of the Copyright Act\nand fulfils the purpose test.\n\n## II. FAIRNESS TEST\n&#8211; Now, I will examine the second aspect, i.e. whether the acts of Open AI\ncan be considered \u201cfair dealing\u201d under Section 52 of the Copyright Act.\n&#8211; The term \u201cfair dealing\u201d has not been defined in the Copyright Act.\nTherefore, it would be useful to examine how Courts in India have interpreted\nthe term \u201cfair dealing\u201d.\n&#8211; Open AI has placed reliance on Narendera Publishing (supra). In the\nsaid judgment, the plaintiff was a well-known publisher of academic books,\nwho filed a suit against the defendants, who were selling guide books, which\nreproduced excerpts from the plaintiff\u2019s books.\n&#8211; In Narendera Publishing (supra), the Court placed partial reliance on\n\n## 78\nEconomic Times, &#8216;I am Sophie&#8217;: 17-year-old from a small UP city\u2019s Hindi-medium school builds AI robot\nteacher   that   can   teach   any   subject,   The   Economic   Times   (Nov.   29,   2025),\nhttps:\/\/economictimes.indiatimes.com\/news\/india\/i-am-sophie-class-12-student-from-a-small-up-citys-\nhindi-medium-school-builds-ai-robot-teacher-that-can-teach-any-\nsubject\/articleshow\/125653059.cms?from=mdr.\n\n\n\nCS(COMM) 1028\/2024     Page 116 of 135\n\nthe four-factor test\n## 79\nwhich is a part of the Copyright Act of 1976 of the United\nStates of America. The Court placed emphasis on the transformative character\nof the use i.e. whether the purpose of the use is substantially different from\nthe purpose served by the prior work. It was also observed that if the use is\ntransformative, it will not serve as a market substitute of the original. Hence,\nsuch use will not affect the market share of the original work. Based on the\nabove,  the Court held  that  the  defendant\u2019s  work  can  be  said  to  be\n\u2018transformative\u2019 and hence, entitled to defence under Section 52(1)(a)(ii) of\nthe Copyright Act. The relevant paragraphs dealing with this aspect from\nNarendera Publishing (supra) are set out below:\n\n\u201c34.  One  crucial  test,  of  the  four-factor  test,  as  developed  by  the\nAmerican courts, is the transformative character of the use. The Courts\nshould in cases like the present ask whether the purpose served by the\nsubsequent (or infringing) work is substantially different (or is the\nsame) from the purpose served by the prior work. The subsequent\nwork must be different in character; it must not be a mere substitute,\nin that, it not sufficient that only superficial changes are made, the\nbasic character remaining the same, to be called transformative. This\ndetermination, according to the Court is closely knit with the other three\nfactors, and therefore, central to the determination of fair use. If the\nwork is transformative, then it might not matter that the copying is\nwhole or substantial. Again, if it is transformative, it may not act as a\n\n## 79\nSection 107 &#8211; Notwithstanding  the  provisions  of  sections  106  and  106A,  the  fair  use  of  a  copyrighted\nwork, including such use by reproduction in copies or phonorecords or by any other means specified by that\nsection,  for  purposes such  as criticism,  comment,  news  reporting,  teaching (including multiple copies for\nclassroom use), scholarship, or research, is not an infringement of copyright. In determining whether the use\nmade of a work in any particular case is a fair use the factors to be considered shall include\u2014\n(1)  the  purpose  and  character  of  the  use, including whether  such  use  is  of  a  commercial  nature  or  is  for\nnonprofit educational purposes;\n(2) the nature of the copyrighted work;\n(3) the amount and substantiality of the portion used in relation to the copyrighted work as a whole; and\n(4) the effect of the use upon the potential market for or value of the copyrighted work.\nThe  fact  that  a  work  is  unpublished  shall  not  itself  bar  a  finding of  fair  use  if  such  finding  is  made  upon\nconsideration of all the above factors.\n\n\n\n\nCS(COMM) 1028\/2024     Page 117 of 135\n\nmarket substitute and consequently, will not affect the market share\nof the prior work.\u201d\n[Emphasis supplied]\n\n&#8211; The judgment in Narendera Publishing (supra) was taken up in appeal\nby the plaintiffs before the Division Bench. The Division Bench\n## 80\ndecided the\nsaid appeal along with another appeal dealing with Section 52(1)(h) [now\nSection 52(1)(i)]. The Division Bench re-emphasized that the purpose of the\nCopyright Act was to balance the rights of the author on one hand and the\nsociety on the other. After referring to the judgments of US Courts dealing\nwith the aspect of fair use, the Division Bench held that the guidebooks used\nby the defendant in B.D. Bhandari (supra), are in the nature of \u2018derivative\nworks\u2019.  The  guide  books  have  a  different  purpose  or  character  than  the\noriginal work of the plaintiffs\/appellants.  Hence, it was held that there was\nno infringement. On this aspect, the observations of the Division Bench are\nset out below:\n\n\u201c78. It follows from the above discussion that when a published work is\nprescribed as text book, a \u2018guide\u2019 can be published provided it fulfils\nthe test of a \u2018derivative work\u2019. At the same time, this guide book has to\nbe materially different from the textbook. Guide book should not be\nverbatim reproduction of the text book. If a guide book reproduces the\noriginal work to a great measure without its contribution, it may amount\nto  plagiarism and may infringe the copyright law of  the person in\noriginal work. On the other hand, if guide book written in a different\nformat, albeit, dealing with the subject matter which is contained in\nthe original book, with a purpose to help, assist and support the\nstudents, the problems given in the text book, then it would be a work\ndifferent from the original work. The guiding principle is to see as to\nwhether such guide book provides explanation and\/or step by step\nprocess for reaching the answer, detailed analysis of any problem with\n\n## 80\nSyndicate of the Press of the University of Cambridge v. B.D. Bhandari, 2011 SCC OnLine Del 3215.\n\n\n\nCS(COMM) 1028\/2024     Page 118 of 135\n\nthe objective of making the task simpler in understanding the subject\nprovided  in  the  textbook.  In  writing  this  guide  book,  if  the\ncontribution of aforesaid nature is made, it may be treated as a work\nwhich is somewhat different in nature. This is the crucial test which\nis to be applied and has been characterized as the transformative\ncharacter of the use by the American Courts. Therefore, the task\nwould be to ascertain as to whether the purpose served by the guide\nbook  is  substantially  different  from  the  purpose  served  by  the\ntextbook. If the guide book is different in character and not a mere\nsubstitute  of  the  original  work\/textbook,  it  would  be  treated  as\ntransformative. However, the character must be substantially different\nand it is not sufficient that superficial changes are made with basic\ncharacter of the textbook creeping in the guide book. In that eventuality,\nit would  not  qualify as a work of  transformative character. If this\nderivative  (sic)  work  in  the  guidebook  has  assumed  different\ncharacter, it would not amount to infringement of the original work,\nthough it would have been drawn, to certain extent, from the original\nwork. And, in that sense, it would amount to \u2018fair use\u2019 as described in\n## Amazon. Com. Inc. (supra).\n## ***   ***   ***\n&#8211; We have already formulated the test for ascertaining as to whether\nin a given situation publication of a guide book would be violative of a\ncopyright in a textbook in detail while dealing with RFA (OS) No. 21 of\n&#8211; It is not necessary to repeat the same. Suffice is to state that it\nwill be permissible to publish a guide book if it qualifies the proper\ndescription of a guide book and becomes a derivative work of the\nauthor who has produced such guide book.\u201d\n[Emphasis supplied]\n\n&#8211; Based on the aforesaid findings, the judgment of the Single Bench in\nNarendera  Publishing (supra)  was  upheld  by  the  Division  Bench.  The\nDivision Bench in B.D. Bhandari (supra) used the expression \u2018derivative\nwork\u2019 as opposed to the \u2018transformative work\u2019 used by Single Bench.\n&#8211; Mr.  Ramanujan, Amicus  Curiae, submits  that  the  judgment  in\nNarendera Publishing (supra) applied the \u2018transformative\u2019 test in the context\nof \u201creview\u201d under Section 52(1)(a)(ii) of the Copyright Act and not in the\n\n\n\nCS(COMM) 1028\/2024     Page 119 of 135\n\ncontext  of  Section  52(1)(a)(i).  In B.D. Bhandari (supra),  the  test  finally\nadopted under Section 52(1)(h) was whether a guidebook is a \u2018derivative\nwork\u2019 (used synonymously for \u2018transformative work\u2019). In paragraph 119 of\nthe Division Bench judgment in B.D. Bhandari (supra), the test of derivative\nwork was extended to Section 52(1)(a)(ii) also.\n&#8211; Even though Division Bench judgment in B.D. Bhandari (supra) is in\nthe context of Section 52(1)(a)(ii) of the Copyright Act, there is no reason why\nthe fair dealing principles discussed therein cannot be applied in the context\nof Section 52(1)(a)(i).\n&#8211; In Rameshwari Photocopy Services (DB) (supra), the plaintiffs, who\nwere  major  publishing  houses,  claimed  that  the  defendant,  a  photocopy\nservices shop situated within the precincts of Delhi School of Economics\n(University of Delhi), was photocopying pages from the books published by\nthe plaintiffs. In the facts of the said case, the Court held that the defendant\nwas entitled to defence under Section 52(1)(h) as it existed then [now Section\n52(1)(i)]. The Division Bench in Rameshwari Photocopy Services (supra)\nalso held that the concept of \u201cfair dealing\u201d would not be applicable in cases\ninvolving Section 52(1)(i). In the Indian context, the concept of \u201cfair dealing\u201d\nis only a part of Section 52(1)(a) of the Copyright Act and not the remaining\nprovisions of Section 52.\n&#8211; The Division Bench specifically noted that the four-factor test is a part\nof the Copyright Act in United States of America (Section 107 of the US\nCopyright Act). Therefore, it was held that the decisions of the Courts in US,\napplying the aforesaid test, would have no persuasive value in the Indian\ncontext.\n&#8211; In Blackwood v. A.N.  Parasuraman (supra),  cited  by  Mr. Adarsh\n\n\n\nCS(COMM) 1028\/2024     Page 120 of 135\n\nRamanujan, Madras High Court adopted a test for determining fair dealing\nbased on quantity and quality of the work that has been reproduced.\n&#8211; In ESPN Star Sports v. Global Broadcast News\n## 81\n, a Coordinate Bench\nof  this Court took  note  of  the  test  laid  down  by  the  English Courts.  In\nparagraph 34, the Court observed that the following three factors have to be\nconsidered for fair dealing:\n\n\u201c34. The test indicated in Ashdown was based on the Textbook \u201cThe\nModern Law of Copyright and Designs (3rd edn, 2000)\u201d by Laddie,\nPrescott and Vitoria. The court held that:\n\u201cthe authors suggest that the success or failure of the defence\ndepends on three factors : (1) whether the alleged fair dealing is\nin commercial competition with the owner&#8217;s exploitation of work,\n(2) whether the work has already been published or otherwise\nexposed to the public and (3) the amount and importance of the\nwork which has been taken.\u201d \u201d\n&#8211; In Super Cassettes v. Chintamani Rao (supra), Coordinate Bench of\nthis Court observed that when assessing fair dealing, the Court must consider\nproportionality and whether there is any intellectual input or any original\nmental exercise undertaken by the defendant.\n&#8211; The Division Bench in India TV Independent News Service v. Yashraj\nFilms (supra), made an observation to the effect that \u201ceven in India, Fair Use\nis determined on the same four factors\u201d i.e. the four-factor test used in US.\nHowever, in the said judgment, the Division Bench did not apply the four-\nfactor test but denied interim injunction to the plaintiff by applying the \u2018de-\nminimis\u2019 principle.\n&#8211; From the abovesaid analysis of judgments of the Indian Courts on the\n\n## 81\n2008 SCC OnLine Del 1766.\n\n\n\nCS(COMM) 1028\/2024     Page 121 of 135\n\naspect of fair dealing, the position that emerges is that different Courts have\nadopted different tests according to the facts and circumstances of the case\nbefore them. There is no single test which has been adopted uniformly or\nconsistently  by  the  Indian Courts. There  is  broad  consensus  amongst  the\ncounsel in the present case also that the US four-factor test is not applicable\nin India and there is no single uniform test for examining whether an act under\nSection 52(1)(a) is fair dealing or not.\n&#8211; In light of the aforesaid, this Court shall attempt to formulate factors to\ndetermine \u201cfair dealing\u201d which are suitable in the facts and circumstances of\nthe present case. Whether use by Open AI amounts to fair dealing or not is\nultimately a question of fact, degree and the overall impression carried by the\n## Court.\n&#8211; As  discussed  above,  Section  52  of  the  Copyright Act  balances  the\nauthor\u2019s right and the public interest. Therefore, Court has to consider if the\ntraining of LLMs by using the original literary works of ANI violates ANI\u2019s\nrights. To determine fairness, it has to be examined whether the aforesaid use\nby  Open  AI  of  ANI\u2019s  works  prejudice  interests  of  ANI,  resulting  in\ncompetition,  actual  or  potential  damage  to  ANI\u2019s  legitimate  commercial\ninterests. Simultaneously, consideration must be given to the extent to which\nthe activities of Open AI serve broader public interests, including scientific\nresearch,  technological  innovation,  access  to  information,  and  the\ndissemination of knowledge. Thus, in the facts and circumstances of the case,\nin  my  opinion,  the  factors  relevant  for  determining  fair  dealing  can  be\nsummarised as under:\na. Whether Open AI\u2019s use of ANI\u2019s original literary works is limited to\ntraining its LLMs underlying ChatGPT?\n\n\n\nCS(COMM) 1028\/2024     Page 122 of 135\n\nb. Whether the usage of ANI\u2019s literary works by Open AI would result in\neconomic competition and would prejudice the legitimate interests of\nANI, thereby causing actual or potential damage to ANI?\nc. Whether the functions performed by Open AI through ChatGPT serve\nthe overall public interest?\n&#8211; The first two factors examine how Open AI\u2019s use of ANI\u2019s works does\nnot affect the author\u2019s rights, and the third factor examines whether functions\nperformed by Open AI serve public interest.\n&#8211; The aforesaid factors are broadly in line with the principles emerging\nfrom Article 9 of Berne Convention for the Protection of Literary and\nArtistic Works (1886) (hereinafter  \u2018Berne  Convention\u2019). Article  9  of  the\nBerne Convention which discusses the rights available to a copyright holder\nand its possible exceptions is reproduced below:\n\nArticle 9 Right of Reproduction:\n&#8211; Generally; 2. Possible exceptions; 3. Sound and visual recordings\n(1) Authors of literary and artistic works protected by this Convention\nshall have the exclusive right of authorizing the reproduction of these\nworks, in any manner or form.\n(2) It shall be a matter for legislation in the countries of the Union to\npermit  the reproduction  of  such  works in  certain  special  cases,\nprovided  that  such reproduction  does  not  conflict  with  a  normal\nexploitation of the work and does not unreasonably prejudice the\nlegitimate interests of the author.\n(3) Any sound or visual recording shall be considered as a reproduction\nfor the purposes of this Convention.\n[Emphasis supplied]\n\n&#8211; The Statements of Objects of the Copyright (Amendment) Act, 1983\n## 82\n\n\n## 82\nStatement of Objects and Reasons of Amendment Act 23 of 1983.\u2014India is a member of the two\nInternational conventions on copyright, namely the Berne Convention for the Protection of Literary and\n\n\n\nCS(COMM) 1028\/2024     Page 123 of 135\n\nprovides that the Copyright Act is in conformity with the Berne Convention.\n&#8211; Now, I proceed to examine the facts and circumstances of the present\ncase in light of the aforesaid factors.\n\na. Whether Open AI\u2019s use of ANI\u2019s original literary works is limited to\ntraining its LLMs underlying ChatGPT?\n&#8211; It has already been observed in the findings in respect of Issue No.2\nthat  ANI  has  not  been  able  to  establish  memorization  of  data  or  its\nregurgitation, or communication of the exact copies of ANI\u2019s works to the\npublic.\n&#8211; The  documents  filed  by  Open AI  (document  nos.  8  and  9  of  the\ndocuments filed by defendant along with the reply to I.A. 43500 of 2024)\ndemonstrate that the LLMs underlying ChatGPT are not trained to reproduce\nor  communicate  the  training  material  to  the  public.  Instead,  the  training\nmaterial enables it to generate new responses and perform novel tasks when\npresented with novel prompts.\n&#8211; This Court has not been given any instance where Open AI has used the\nliterary works of ANI for any purposes other than for training. Hence, in my\nopinion, it would be safe to conclude that the storage by Open AI is only for\nthe purposes of training Open AI\u2019s LLMs underlying ChatGPT.\n\n\nArtistic Works (1948 Brussels Text) and the Universal Copyright Convention (1952). The Copyright Act,\n1957, conforms to the provisions of these two conventions.\n\n\n\nCS(COMM) 1028\/2024     Page 124 of 135\n\nb. Whether the usage of ANI\u2019s literary works by Open AI would result\nin economic competition and would prejudice the legitimate interests\nof ANI, thereby causing actual or potential damage to ANI?\n&#8211; ANI  is  a  news  agency  that  is  in  the  business  of  collecting  and\ndistributing news. ANI has engaged a team of professionals to collectively\nproduce original news content, which is published on ANI\u2019s syndicated news\nfeed. The subscribers of ANI, which include news organisations, can access\nthe content posted on ANI\u2019s syndicated news feed. A non-exclusive license is\ngranted  to ANI\u2019s  subscribers,  who  can  use ANI\u2019s  works  for  commercial\npurposes, specifically excluding any right of further syndication of ANI\u2019s\nworks.  Apart  from  the  revenue  generated  from  these  subscription  and\nlicensing  agreements,  ANI  earns  by  publishing  the  news  articles  on  its\nwebsite, YouTube channel and other social media accounts. The user traffic\non its websites and social media channels is directly correlated to revenue\nraised through advertising by ANI.\n## 83\n\n&#8211; On the other hand, Open AI\u2019s ChatGPT is an application, which allows\nthe users to interact with LLMs. LLMs serve as tools for content creation,\nresearch,  language  translation,  learning  and  education,  summarising,\nparaphrasing,  creation  of  images  based  on  text  description.  ChatGPT is\navailable for free to all users. There is also an option for users to sign up for a\npremium monthly subscription that makes additional features and capabilities\navailable.  Open  AI  also  offers  an  enterprise  version  for  businesses  to\npurchase. The API (Application Programming Interface) version of Open AI\nallows the developers to integrate the benefits of Open AI\u2019s models into their\n\n## 83\n\u00b6\u00b6 9-15 of the plaint.\n\n\n\nCS(COMM) 1028\/2024     Page 125 of 135\n\nown application. The developers pay to access the API.\n## 84\n\n&#8211; As is evident from aforesaid, the functions performed by ChatGPT are\nquite  distinct  from  the  functions  performed  by  ANI.  ChatGPT  has\nmultifarious functions, whereas business of ANI is limited to news reporting\nand syndication of news. In the context of news, ChatGPT only provides a\nsummary or snippets of the news articles or reports carried by ANI along with\na reference to its website. As held in Issue No.2, the expressions used are not\nthe same as used by ANI. Even when a user searches for daily news, the\nresponses  given  by  ChatGPT  are  similar  to  that  of  an AI-enabled  search\nengine, using RAG technique to provide multiple sources for the snippets of\nthe news content. In sum, the purpose and character of use of copyrighted\nworks by Open AI is fundamentally different from that of the news articles of\nANI.  Hence,  it  cannot  be  said  that  responses  produced  by  ChatGPT  are\nsubstitutes for news articles published by ANI.\n&#8211; This view also finds support from the judgment of the United States\nDistrict Court for  Northern  District  of  California  in Bartz v. Anthropic\n(supra), wherein it has been observed that use of copyright to train LLMs was\nessentially transformative:\n\n\u201c      The  copies  used  to  train  specific  LLMs  did  not  and  will  not\ndisplace demand for copies of Authors\u2019 works, or not in the way that\ncounts under the Copyright Act.\n\nInstead,  Authors  contend  generically  that  training  LLMs  will\nresult in an explosion of works competing with their works &#8211; such as by\ncreating alternative summaries of factual events, alternative examples\nof  compelling  writing  about  fictional  events, and  so  on.  This  order\nassumes  that  is  so  (Opp.  22-23  (citing,  e.g.,  Opp.  Exh.  38)). But\n\n## 84\nDocument 8 of the documents filed by the defendant along with reply to I.A. No. 45300 of 2024.\n\n\n\nCS(COMM) 1028\/2024     Page 126 of 135\n\nAuthors\u2019 complaint is no different than it would be if they complained\nthat training schoolchildren to write well would result in an explosion\nof  competing  works.  This  is  not  the  kind  of  competitive  or  creative\ndisplacement  that  concerns  the  Copyright  Act. The  Act  seeks  to\nadvance original works of authorship, not to protect authors against\ncompetition.\n## ***   ***   ***\nIn short, the purpose and character of using copyrighted works to\ntrain LLMs to generate new text was quintessentially transformative.\nLike any reader aspiring to be a writer, Anthropic\u2019s LLMs trained\nupon works not to race ahead and replicate or supplant them \u2014 but\nto turn a hard corner and create something different. If this training\nprocess  reasonably  required  making  copies  within  the  LLM  or\notherwise, those copies were engaged in a transformative use.\u201d\n\n[Emphasis supplied]\n\n&#8211; Similar observations were also made by the United States District Court\nfor Northern District of California in Kadrey v. Meta Platforms, (supra) i.e.\nthe purpose should be transformative:\n\u201cThis factor favors Meta. There is no serious question that Meta\u2019s use\nof  the  plaintiffs\u2019 books  had  a \u201cfurther  purpose\u201d and \u201cdifferent\ncharacter\u201d than the books that it was highly transformative. The\npurpose of Meta\u2019s copying was to train its LLMs, which are innovative\ntools that can be used to generate diverse text and perform a wide\nrange of functions. Cf. Oracle, 593 US. at 30 (transformative to use\ncopyrighted computer code \u201cto create a new platform that could be\nreadily used by programmers\u201d) Users can ask Llama to edit an email\nthey have written, translate an excerpt from or into a foreign language,\nwrite a skit based on a hypothetical scenario. or do any number of other\ntasks. The purpose of the plaintiffs books, by contrast, is to be read for\nentertainment or education.\n\nThe plaintiffs do not meaningfully disagree about Llama\u2019s purpose.\nTo  the  contrary,  they  acknowledge  that  LLMs  have \u201cend  uses\u201d\nincluding  serving \u201cas  a  personal  tutor,\u201d assisting \u201cwith  creative\nideation,\u201d and helping users \u201cgenerate business reports.\u201d And several\nof the plaintiffs testified to using LLMs for various purposes, all distinct\n\n\n\nCS(COMM) 1028\/2024     Page 127 of 135\n\nfrom creating or reading in expressive work like a novel or biography\nfor  instance,  to  find  recipes,  get  tax  or  medical  advice,  translate\ndocuments, or conduct research. All of these functions are different\nfrom  the  use  to  which  the  plaintiffs  books  are  generally  put.  So\ncopying the books to develop a tool that can perform those functions\nis  a  use  with  a  different  purpose  and  character  than  the  books\nthemselves.\u201d\n[Emphasis supplied]\n\n&#8211; In Authors Guild v. Google\n## 85\n, the authors of published books filed a\ncopyright  infringement  action  against  Google  claiming  that  Google  made\ndigital copies of the books and allowed the public to search the texts of the\ndigitally  copied  books  and  see  displays  of  snippets  of  the  text,  which\namounted to infringement of the copyright of the authors. The United States\nCourt of Appeals held that the more transformative the new work, the lesser\nwill  be  the  significance  of  commercialism.  In  the  said  case,  Google  was\ndigitizing  copyright  protected  works  of  authors.  The Court held  that  the\npurpose of copying is highly transformative as the public display of text is\nlimited and the activities of Google are not in the nature of market substitution\nof  the  original  books.  The  mere  fact  that  the  activities  of  Google  are\ncommercial in  nature  and  profit-motivated  would  not  deny  Google  the\ndefence of fair use.\n&#8211; Except for bare averments, nothing has been placed on record on behalf\nof ANI to show that they have lost any market share or there has been a\nreduction in subscription revenues due to operations of Open AI.\n&#8211; Hence, in my prima facie opinion, the use of ANI\u2019s works by Open AI\nwould not result in market substitution of ANI\u2019s works and consequently, will\n\n## 85\n## 804 F.3d 202 (2nd Cir. 2015).\n\n\n\nCS(COMM) 1028\/2024     Page 128 of 135\n\nnot affect its market share or cause actual or potential damage to ANI.\n\nc. Whether the functions performed by Open AI through ChatGPT\nserve the overall public interest?\n&#8211; The  public  benefits  flowing  from  such  trained  LLMs  underlying\nChatGPT are considerable. They are designed to assist users in analysing and\ngenerating  text,  improving  access  to  information,  supporting  education,\nassisting  scientific  research,  facilitating  software  development,  enabling\ntranslation  and  communication,  and  creating  tools  for  persons  with\ndisabilities. The technology is also capable of supporting research, learning,\nand knowledge dissemination at scale, thereby advancing science and the\nuseful arts in a broader societal sense.\n&#8211; Training  LLMs  underlying  ChatGPT  therefore  contributes  to\nadvancing scientific knowledge, developing innovative computational tools,\ndisseminating information, promoting education, enhancing accessibility, and\nadvancing artificial intelligence research. These objectives resonate with the\nbroader constitutional and statutory commitment to advancing knowledge and\nlearning. While the rights of copyright owners remain important, the societal\nbenefits  arising  from  scientific  and  technological  research  constitute a\nrelevant consideration in assessing the fairness of a dealing.\n&#8211; Therefore, the factor of public interest also stands established.\n&#8211; Thus, on a prima facie view, all the factors for establishing the aspect\nof fair dealing stand satisfied in the present case and the fairness test stands\nfulfilled.\n\n\n\n\nCS(COMM) 1028\/2024     Page 129 of 135\n\n## FINDINGS\n&#8211; In light of the discussion above, both the purpose test as well as the\nfairness test under Section 52(1)(a) stand fulfilled. Hence, in my prima facie\nview, Open AI\u2019s acts of storage of the literary works of ANI for the training\nof its LLMs would fall under Section 52(1)(a) of the Copyright Act and hence,\nwould not amount to infringement.\n\n## BALANCE OF CONVENIENCE AND IRREPARABLE INJURY\n## SUBMISSIONS\n\nSubmissions on behalf of the Plaintiff (ANI)\n&#8211; Mr.  Sidhant  Kumar,  counsel  appearing  on  behalf  of ANI  made  the\nfollowing submissions:\n257.1. Open AI, by offering direct quotes and summaries of ANI\u2019s work, is\ndiverting traffic from ANI\u2019s website to ChatGPT thereby reducing the revenue\ngenerated  by  ANI  through  advertisements.  Open  AI\u2019s  unlawful  use  and\nstorage  of ANI\u2019s  works  curtails ANI\u2019s  commercial  avenues  including  the\nability to use AI to provide services including high level summaries of its\nworks.\n257.2. Open  AI  has  entered  into  licensing  arrangements  and  strategic\npartnerships  with  similarly-placed  news  agencies  with  articles,  news  and\nother works in the training and response generation process. However, it has\nconsciously opted not to compensate ANI for using its data. Therefore, Open\nAI is unjustly enriching itself at the expense of ANI.\n\nSubmissions on behalf of the Defendant (OPEN AI)\n&#8211; Mr. Amit Sibal, senior counsel appearing on behalf of Open AI made\n\n\n\nCS(COMM) 1028\/2024     Page 130 of 135\n\nthe following submissions:\n258.1. The interim relief sought by ANI to delete the \u201cplaintiff\u2019s works held\nand stored in whatever form\u201d by Open AI would amount to grant of the final\nrelief  sought  in  the suit. This  would  also  be inconsistent  with  Open AI\u2019s\nobligation  under  the  laws  of  United  States  of America  to  preserve  such\ntraining data.\n258.2. ANI itself had offered Open AI a license to all its digital media video,\nimagery, photographic and\/or news content owned or within the control of\nANI for a fee of USD 7.5 million in terms of ANI\u2019s letter dated 3\nrd\n## October\n&#8211; Therefore, it is clear that the claim of ANI is quantifiable and ANI can\nbe compensated in monetary terms if ANI succeeds in the suit. On the other\nhand,  Open AI  cannot  be  compensated  in  monetary  terms  if  an  interim\ninjunction is granted against Open AI, even if Open AI succeeds in the suit.\n258.3. The  AI  platform  of  Open  AI, i.e.  ChatGPT, has  transformed\ninformation in new ways and is being used all around the world for a variety\nof beneficial purposes. Any injunction granted would be detrimental to public\ninterest.\n258.4. Without prejudice to their rights, Open AI has stopped ANI\u2019s website\nboth for the purposes of training as well as RAG.\n\nIntervenors supporting OPEN AI\ni. Submissions on behalf of the Intervenor &#8211; Broadband India Forum (BIF)\n## (I.A. 10766\/2025)\n&#8211; Mr.  Arvind  P.  Datar, senior  counsel appearing  on  behalf  of  the\nintervenor in I.A. 10766\/2025 made the following submissions:\n259.1. Any injunction granted will have cascading effects for several LLM\n\n\n\nCS(COMM) 1028\/2024     Page 131 of 135\n\nplatforms as it would curtail access to information for LLM developers and\nthe public. It would also require an LLM to take multiple licenses with each\nnews  agency,  individual  newspaper,  magazine  etc.  Hence,  when ANI  has\nchosen not to engage \u2018paywalls\u2019 or \u2018crawler blockers\u2019, no injunction can be\ngranted to access information that is freely available on the web.\n259.2. Hindering access to publicly available data would degrade LLM\u2019s\nperformance and stifle information as developers would struggle to improve\ntheir AI models.\n259.3. ANI  has  claimed  damages  in  the  suit  and  hence,  no  interlocutory\ninjunction should be granted in favour of ANI. The issues raised by ANI\nwould require large amount of evidence in the trial. Therefore, no injunction\nshould be granted in the absence of a trial.\n\nSubmissions of Professor Arul George Scaria, Amicus Curiae\n&#8211; Prof. Scaria made the following submissions:\n260.1. In the event the Court finds that the Open AI is required to take a\nlicense from ANI, it would be detrimental to the growth of LLMs and more\nparticularly, LLMs being sought to be developed in India.\n260.2. While granting an interim injunction, the Court should consider the\ntwo competing interests, i.e. protecting the interest of the authors and the\nbroader public interest in developing LLMs in India.\n260.3. It  is  an undisputed  fact  that  at least  some  users  use  ChatGPT  for\ngathering information about events and facts. Therefore, the Courts should\ntake a broader view so as to enable fulfilment of this broader purpose.\n\n\n\n\n\n\nCS(COMM) 1028\/2024     Page 132 of 135\n\nSubmissions of Mr. Adarsh Ramanujan, Amicus Curiae\n&#8211; Mr. Ramanujan made the following submissions:\n261.1. ChatGPT is currently being used by millions of users in India and a\nblanket injunction order passed by this Court would adversely affect the said\nusers.\n261.2. Insofar as ANI seeks a direction that Open AI should obtain license\nfrom ANI in respect of its works, it is demonstrative of the fact that ANI can\nbe compensated in monetary terms.\n261.3. ANI is in a position to block its websites to prevent Open AI from\nscraping data therefrom for training of its LLM as also for the purposes of its\nsearch functionality.\n## DISCUSSION AND ANALYSIS\n&#8211; It is an admitted position that ANI has the ability to block its website\nvis-\u00e0-vis any third-party including Open AI. The opting-out option is available\nto ANI for blocking the third-party web crawlers from copying their data as\nwell as from scraping their website for the search function\/RAG. Despite\nhaving an option of opt-out, evidently ANI has not exercised the same. In fact,\nit has been stated on behalf of Open AI that it has internally blocked ANI\u2019s\nwebsite from its web crawlers or bots for the purposes of scraping of data.\n## 86\n\nDuring the course of oral submissions Open AI has also submitted that Open\nAI itself has blocked ANI\u2019s website from \u2018ChatGPT search function\/RAG\u2019.\n&#8211; ANI contends that there are various paid subscribers of ANI, whose\nwebsites carry the copyrighted data of ANI and Open AI can still access and\ncopy ANI\u2019s copyrighted data from the aforesaid websites. Even if that be so,\n\n## 86\nDocument 21 of the plaintiff\u2019s documents filed along with the plaint.\n\n\n\nCS(COMM) 1028\/2024     Page 133 of 135\n\nit would be open for the subscribers of ANI also to block their websites to\nOpen AI. Clearly, if they have chosen not to do so, it could be possibly for\ntheir own commercial reasons.\n&#8211; There is no averment in the plaint nor has any material been placed on\nrecord to show that the activities of Open AI have resulted in ANI losing\nsubscribers or suffering a loss from its news syndication business.\n&#8211; ANI itself has offered a license to Open AI for its content for a fee of\nUSD 7.5 million vide its communication dated 3\nrd\n## October 2024.\n## 87\n## Therefore,\nthe claim of ANI is quantifiable and ANI can be compensated in monetary\nterms, if ANI succeeds in the present case. On the other hand, if there is an\ninjunction granted against Open AI, it would have a significant effect on its\nfunctioning, which may not be possible to be compensated in monetary terms.\n&#8211; There is yet another aspect of public interest for not granting an interim\ninjunction  in  favour  of ANI  and  against  Open AI. As  discussed  earlier,\nArtificial  Intelligence  (AI)  including  generative  AI  has brought  about a\ntransformational change in the lives of the people all over the world, including\nIndia. It has completely revolutionized the manner in which people seek and\nobtain information.\n&#8211; As noted earlier, AI and its applications are being used beneficially in\nseveral  sectors,  such  as education, healthcare,  financial  support  sector,\nagriculture and for providing other skill development resources. [Ref. AI for\nInclusive Societal Development, October 2025 &#8211; Niti Aayog paper at page\n7389 of Pleadings Vol. 6]\n&#8211; The key to success of an AI model is to access the information in public\n\n## 87\nDocument 24 of the plaintiff\u2019s documents filed along with the plaint.\n\n\n\nCS(COMM) 1028\/2024     Page 134 of 135\n\ndomain. The development of LLMs and their success depends on availability\nof data. It would be economically unviable to develop an LLM if training of\nan LLM would require licenses from multiple sources.\n&#8211; Any interim injunction granted at this stage would, in my opinion, be\ndetrimental to the growth of AI and more particularly, to the LLMs being\ndeveloped in India. It would also have adverse impact on public interest,\nincluding millions of users of ChatGPT in India, many of whom would not be\npaid subscribers. Public interest is an important aspect or the fourth factor that\nthe Courts  have  to  consider  while  granting  interim  injunction,  more\nparticularly in intellectual property cases. [Pls. see. Zydus Lifesciences v.\nE.R. Squibb\n## 88\n; F-Hoffman-La Roche AG v. NATCO Pharma\n## 89\n## ; Astrazeneca\nAB and Ors. v. Intas Pharmaceuticals Limited\n## 90\n## ].\n\n## FINDINGS\n&#8211; In  light  of  the  aforesaid  analysis,  on  the  touchstone  of balance of\nconvenience as well as irreparable loss and injury, this Court is not inclined\nto grant an interim injunction this stage.\n## CONCLUSION\n&#8211; Based on the discussion above, I am of the prima facie view that Open\nAI\u2019s act of storing ANI\u2019s original literary works for training LLMs underlying\nChatGPT falls under Section 52(1)(a) of the Copyright Act and therefore, does\nnot amount to infringement under Section 51 of the Copyright Act. I am also\nof the prima facie view that the outputs generated by ChatGPT using RAG\n\n## 88\nFAO(OS) (COMM) 120\/2025, decided on 12\nth\n## January 2026.\n## 89\nCS(COMM) 567\/2024, decided on 24\nth\n## March, 2025.\n## 90\n## 2020:DHC:3125.\n\n\n\nCS(COMM) 1028\/2024     Page 135 of 135\n\ntechnique does not amount to infringement under Section 51 of the Copyright\nAct since the outputs generated by Open AI were not substantially similar to\nANI\u2019s original literary works. Further, ANI has failed to satisfy this Court that\nany  memorization  and  regurgitation  of ANI\u2019s  original  literary  works  has\nhappened through the responses generated by ChatGPT.\n&#8211; In view of the discussion above, ANI has failed to make out a prima\nfacie case for grant of interim injunction. Balance of convenience is also\nagainst grant of interim injunction. Irreparable injury would be caused not\nonly to Open AI but also to the public at large, if an interim injunction is\ngranted in favour of ANI at this stage.\n&#8211; Accordingly, I.A. 45300\/2024 is dismissed.\n&#8211; Needless to say, any observations made herein are only for the purpose\nof adjudication of the aforesaid application and would have no bearing on the\nfinal outcome of the suit.\n&#8211; I  would  like  to  express  my  sincere  appreciation  to  all  the  learned\ncounsel who appeared in the matter as well as the learned Amici Curiae who\nrendered invaluable assistance to the Court in rendering this judgment.\n\n\n\n## AMIT BANSAL\n## (JUDGE)\n\n## JULY 24, 2026\n## Vivek\/-\n\n\n\n<p><\/p>\n","protected":false},"excerpt":{"rendered":"<p>Markdown File: CS(COMM) 1028\/2024 Page 1 of 135 ## * IN THE HIGH COURT OF DELHI AT NEW DELHI % Judgment reserved [&hellip;]<\/p>\n","protected":false},"author":1,"featured_media":0,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"_acf_changed":false,"footnotes":""},"categories":[1],"tags":[],"class_list":["post-632","post","type-post","status-publish","format-standard","hentry","category-uncategorized"],"acf":[],"yoast_head":"<!-- This site is optimized with the Yoast SEO Premium plugin v25.8 (Yoast SEO v27.4) - https:\/\/yoast.com\/product\/yoast-seo-premium-wordpress\/ -->\n<title>ANI Media Pvt. Ltd. v Open AI OPCO LLC - Varun Matlani<\/title>\n<meta name=\"robots\" content=\"index, follow, max-snippet:-1, max-image-preview:large, max-video-preview:-1\" \/>\n<link rel=\"canonical\" href=\"https:\/\/matlani.in\/?p=632\" \/>\n<meta property=\"og:locale\" content=\"en_US\" \/>\n<meta property=\"og:type\" content=\"article\" \/>\n<meta property=\"og:title\" content=\"ANI Media Pvt. Ltd. v Open AI OPCO LLC\" \/>\n<meta property=\"og:description\" content=\"Markdown File: CS(COMM) 1028\/2024 Page 1 of 135 ## * IN THE HIGH COURT OF DELHI AT NEW DELHI % Judgment reserved [&hellip;]\" \/>\n<meta property=\"og:url\" content=\"https:\/\/matlani.in\/?p=632\" \/>\n<meta property=\"og:site_name\" content=\"Varun Matlani\" \/>\n<meta property=\"article:published_time\" content=\"2026-07-26T13:22:53+00:00\" \/>\n<meta name=\"author\" content=\"Varun Matlani\" \/>\n<meta name=\"twitter:card\" content=\"summary_large_image\" \/>\n<meta name=\"twitter:label1\" content=\"Written by\" \/>\n\t<meta name=\"twitter:data1\" content=\"Varun Matlani\" \/>\n\t<meta name=\"twitter:label2\" content=\"Est. reading time\" \/>\n\t<meta name=\"twitter:data2\" content=\"1 minute\" \/>\n<script type=\"application\/ld+json\" class=\"yoast-schema-graph\">{\"@context\":\"https:\\\/\\\/schema.org\",\"@graph\":[{\"@type\":\"Article\",\"@id\":\"https:\\\/\\\/matlani.in\\\/?p=632#article\",\"isPartOf\":{\"@id\":\"https:\\\/\\\/matlani.in\\\/?p=632\"},\"author\":{\"name\":\"Varun Matlani\",\"@id\":\"https:\\\/\\\/matlani.in\\\/#\\\/schema\\\/person\\\/cb8a891be76391d2647136a8e7b08be8\"},\"headline\":\"ANI Media Pvt. Ltd. v Open AI OPCO LLC\",\"datePublished\":\"2026-07-26T13:22:53+00:00\",\"mainEntityOfPage\":{\"@id\":\"https:\\\/\\\/matlani.in\\\/?p=632\"},\"wordCount\":37658,\"commentCount\":0,\"inLanguage\":\"en-US\",\"potentialAction\":[{\"@type\":\"CommentAction\",\"name\":\"Comment\",\"target\":[\"https:\\\/\\\/matlani.in\\\/?p=632#respond\"]}]},{\"@type\":\"WebPage\",\"@id\":\"https:\\\/\\\/matlani.in\\\/?p=632\",\"url\":\"https:\\\/\\\/matlani.in\\\/?p=632\",\"name\":\"ANI Media Pvt. Ltd. v Open AI OPCO LLC - Varun Matlani\",\"isPartOf\":{\"@id\":\"https:\\\/\\\/matlani.in\\\/#website\"},\"datePublished\":\"2026-07-26T13:22:53+00:00\",\"author\":{\"@id\":\"https:\\\/\\\/matlani.in\\\/#\\\/schema\\\/person\\\/cb8a891be76391d2647136a8e7b08be8\"},\"breadcrumb\":{\"@id\":\"https:\\\/\\\/matlani.in\\\/?p=632#breadcrumb\"},\"inLanguage\":\"en-US\",\"potentialAction\":[{\"@type\":\"ReadAction\",\"target\":[\"https:\\\/\\\/matlani.in\\\/?p=632\"]}]},{\"@type\":\"BreadcrumbList\",\"@id\":\"https:\\\/\\\/matlani.in\\\/?p=632#breadcrumb\",\"itemListElement\":[{\"@type\":\"ListItem\",\"position\":1,\"name\":\"Home\",\"item\":\"https:\\\/\\\/matlani.in\\\/\"},{\"@type\":\"ListItem\",\"position\":2,\"name\":\"ANI Media Pvt. Ltd. v Open AI OPCO LLC\"}]},{\"@type\":\"WebSite\",\"@id\":\"https:\\\/\\\/matlani.in\\\/#website\",\"url\":\"https:\\\/\\\/matlani.in\\\/\",\"name\":\"Varun Matlani\",\"description\":\"\",\"potentialAction\":[{\"@type\":\"SearchAction\",\"target\":{\"@type\":\"EntryPoint\",\"urlTemplate\":\"https:\\\/\\\/matlani.in\\\/?s={search_term_string}\"},\"query-input\":{\"@type\":\"PropertyValueSpecification\",\"valueRequired\":true,\"valueName\":\"search_term_string\"}}],\"inLanguage\":\"en-US\"},{\"@type\":\"Person\",\"@id\":\"https:\\\/\\\/matlani.in\\\/#\\\/schema\\\/person\\\/cb8a891be76391d2647136a8e7b08be8\",\"name\":\"Varun Matlani\",\"image\":{\"@type\":\"ImageObject\",\"inLanguage\":\"en-US\",\"@id\":\"https:\\\/\\\/secure.gravatar.com\\\/avatar\\\/ad03eb2423d30c16f00b649c50e57a5ae8f667a2dd0799604727481b8b1b96a4?s=96&d=mm&r=g\",\"url\":\"https:\\\/\\\/secure.gravatar.com\\\/avatar\\\/ad03eb2423d30c16f00b649c50e57a5ae8f667a2dd0799604727481b8b1b96a4?s=96&d=mm&r=g\",\"contentUrl\":\"https:\\\/\\\/secure.gravatar.com\\\/avatar\\\/ad03eb2423d30c16f00b649c50e57a5ae8f667a2dd0799604727481b8b1b96a4?s=96&d=mm&r=g\",\"caption\":\"Varun Matlani\"},\"description\":\"Securities Lawyer &amp; Fund Formation Counsel Advocate, Bar Council of Maharashtra &amp; Goa Gujarat National Law University; ICAI Areas of Expertise - Securities Law, Alternative Investment Funds, Direct Tax &amp; Transfer Pricing, Life Sciences &amp; Medical Laws, Insolvency. (In this order) Started-up: BarodaBookStores (2018-20), 2NLU (2020-22), SecMarx AI\",\"sameAs\":[\"https:\\\/\\\/matlani.in\"],\"url\":\"https:\\\/\\\/matlani.in\\\/?author=1\"}]}<\/script>\n<!-- \/ Yoast SEO Premium plugin. -->","yoast_head_json":{"title":"ANI Media Pvt. Ltd. v Open AI OPCO LLC - Varun Matlani","robots":{"index":"index","follow":"follow","max-snippet":"max-snippet:-1","max-image-preview":"max-image-preview:large","max-video-preview":"max-video-preview:-1"},"canonical":"https:\/\/matlani.in\/?p=632","og_locale":"en_US","og_type":"article","og_title":"ANI Media Pvt. Ltd. v Open AI OPCO LLC","og_description":"Markdown File: CS(COMM) 1028\/2024 Page 1 of 135 ## * IN THE HIGH COURT OF DELHI AT NEW DELHI % Judgment reserved [&hellip;]","og_url":"https:\/\/matlani.in\/?p=632","og_site_name":"Varun Matlani","article_published_time":"2026-07-26T13:22:53+00:00","author":"Varun Matlani","twitter_card":"summary_large_image","twitter_misc":{"Written by":"Varun Matlani","Est. reading time":"1 minute"},"schema":{"@context":"https:\/\/schema.org","@graph":[{"@type":"Article","@id":"https:\/\/matlani.in\/?p=632#article","isPartOf":{"@id":"https:\/\/matlani.in\/?p=632"},"author":{"name":"Varun Matlani","@id":"https:\/\/matlani.in\/#\/schema\/person\/cb8a891be76391d2647136a8e7b08be8"},"headline":"ANI Media Pvt. Ltd. v Open AI OPCO LLC","datePublished":"2026-07-26T13:22:53+00:00","mainEntityOfPage":{"@id":"https:\/\/matlani.in\/?p=632"},"wordCount":37658,"commentCount":0,"inLanguage":"en-US","potentialAction":[{"@type":"CommentAction","name":"Comment","target":["https:\/\/matlani.in\/?p=632#respond"]}]},{"@type":"WebPage","@id":"https:\/\/matlani.in\/?p=632","url":"https:\/\/matlani.in\/?p=632","name":"ANI Media Pvt. Ltd. v Open AI OPCO LLC - Varun Matlani","isPartOf":{"@id":"https:\/\/matlani.in\/#website"},"datePublished":"2026-07-26T13:22:53+00:00","author":{"@id":"https:\/\/matlani.in\/#\/schema\/person\/cb8a891be76391d2647136a8e7b08be8"},"breadcrumb":{"@id":"https:\/\/matlani.in\/?p=632#breadcrumb"},"inLanguage":"en-US","potentialAction":[{"@type":"ReadAction","target":["https:\/\/matlani.in\/?p=632"]}]},{"@type":"BreadcrumbList","@id":"https:\/\/matlani.in\/?p=632#breadcrumb","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https:\/\/matlani.in\/"},{"@type":"ListItem","position":2,"name":"ANI Media Pvt. Ltd. v Open AI OPCO LLC"}]},{"@type":"WebSite","@id":"https:\/\/matlani.in\/#website","url":"https:\/\/matlani.in\/","name":"Varun Matlani","description":"","potentialAction":[{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https:\/\/matlani.in\/?s={search_term_string}"},"query-input":{"@type":"PropertyValueSpecification","valueRequired":true,"valueName":"search_term_string"}}],"inLanguage":"en-US"},{"@type":"Person","@id":"https:\/\/matlani.in\/#\/schema\/person\/cb8a891be76391d2647136a8e7b08be8","name":"Varun Matlani","image":{"@type":"ImageObject","inLanguage":"en-US","@id":"https:\/\/secure.gravatar.com\/avatar\/ad03eb2423d30c16f00b649c50e57a5ae8f667a2dd0799604727481b8b1b96a4?s=96&d=mm&r=g","url":"https:\/\/secure.gravatar.com\/avatar\/ad03eb2423d30c16f00b649c50e57a5ae8f667a2dd0799604727481b8b1b96a4?s=96&d=mm&r=g","contentUrl":"https:\/\/secure.gravatar.com\/avatar\/ad03eb2423d30c16f00b649c50e57a5ae8f667a2dd0799604727481b8b1b96a4?s=96&d=mm&r=g","caption":"Varun Matlani"},"description":"Securities Lawyer &amp; Fund Formation Counsel Advocate, Bar Council of Maharashtra &amp; Goa Gujarat National Law University; ICAI Areas of Expertise - Securities Law, Alternative Investment Funds, Direct Tax &amp; Transfer Pricing, Life Sciences &amp; Medical Laws, Insolvency. (In this order) Started-up: BarodaBookStores (2018-20), 2NLU (2020-22), SecMarx AI","sameAs":["https:\/\/matlani.in"],"url":"https:\/\/matlani.in\/?author=1"}]}},"_links":{"self":[{"href":"https:\/\/matlani.in\/index.php?rest_route=\/wp\/v2\/posts\/632","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/matlani.in\/index.php?rest_route=\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/matlani.in\/index.php?rest_route=\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/matlani.in\/index.php?rest_route=\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/matlani.in\/index.php?rest_route=%2Fwp%2Fv2%2Fcomments&post=632"}],"version-history":[{"count":1,"href":"https:\/\/matlani.in\/index.php?rest_route=\/wp\/v2\/posts\/632\/revisions"}],"predecessor-version":[{"id":634,"href":"https:\/\/matlani.in\/index.php?rest_route=\/wp\/v2\/posts\/632\/revisions\/634"}],"wp:attachment":[{"href":"https:\/\/matlani.in\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=632"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/matlani.in\/index.php?rest_route=%2Fwp%2Fv2%2Fcategories&post=632"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/matlani.in\/index.php?rest_route=%2Fwp%2Fv2%2Ftags&post=632"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}