Published On: July 21st 2026
Authored By: R Srikant
Bharati Vidyapeeth New Law College
I. Introduction
Artificial Intelligence (AI) has rapidly transformed the way information is created, processed, and disseminated. Generative AI systems such as ChatGPT, Gemini, and Claude are capable of producing human-like responses by training on massive datasets collected from various online sources. While these technological developments have unlocked significant opportunities for innovation, they have simultaneously raised complex legal questions regarding copyright protection, ownership of content, and the permissible use of copyrighted material for AI training.
The intersection between copyright law and AI has emerged as one of the most significant legal challenges of the digital age. Around the world, authors, publishers, artists, and media organizations have initiated legal proceedings against AI companies, alleging unauthorized use of copyrighted content for training large language models (LLMs). India has now entered this global debate through the landmark litigation between Asian News International (ANI) and OpenAI.
ANI, one of India’s leading news agencies, has alleged that OpenAI utilized its copyrighted news content without authorization to train ChatGPT. OpenAI has denied these allegations and has argued that the use of publicly available information for AI training does not amount to copyright infringement. The dispute has consequently become India’s first major judicial examination of the applicability of the Copyright Act, 1957 to generative AI technologies.
This article examines the legal issues arising from the ANI v. OpenAI litigation, evaluates the competing arguments advanced by both parties, and analyses the broader implications of the dispute for copyright law, technological innovation, and AI regulation in India.
II. Background of the Dispute
The dispute originated when ANI filed a copyright infringement suit before the Delhi High Court in November 2024. ANI alleged that OpenAI had used its copyrighted news reports and journalistic content to train ChatGPT without obtaining a license or authorization. ANI further contended that such use violated its exclusive rights under the Copyright Act, 1957. The litigation has attracted substantial public and legal attention because it represents one of the earliest attempts in India to determine how traditional copyright principles apply to AI systems.[1]
According to ANI, the process of collecting, scraping, storing, and tokenizing copyrighted news content for AI training constitutes reproduction of copyrighted works. ANI argues that the commercial nature of OpenAI’s operations further strengthens the claim that the use of such material cannot be justified under existing copyright exceptions.[2]
OpenAI has strongly contested these allegations. It maintains that its models learn statistical relationships between words rather than storing or reproducing copyrighted expression. The company has argued that AI training involves processing publicly available information in a transformative manner and does not result in direct copying of protected content. OpenAI has also challenged the maintainability of the proceedings on jurisdictional grounds by emphasizing that its model training infrastructure is located outside India.[3]
The litigation has expanded beyond the interests of ANI alone. Various publishers, media organizations, and copyright stakeholders have closely monitored the proceedings, recognizing that the outcome may influence future disputes involving books, music, journalism, and other creative industries.[4]
III. Key Legal Issues Raised by the Litigation
A. Copyright Infringement and AI Training
The primary legal issue concerns whether the use of copyrighted material for training AI models constitutes copyright infringement under Indian law.
Section 14 of the Copyright Act, 1957 grants copyright owners exclusive rights over reproduction and communication of their works. Section 51 further provides that unauthorized exercise of these rights may amount to infringement.
ANI argues that the process of data scraping, storage, and tokenization necessarily involves copying copyrighted content and therefore falls within the scope of reproduction. From this perspective, infringement occurs at the training stage itself, irrespective of whether the final AI output reproduces the original work verbatim.
OpenAI, on the other hand, argues that AI models do not store expressive content in a retrievable form. Instead, they learn patterns, probabilities, and linguistic relationships. Consequently, the company contends that model training should not be equated with traditional reproduction recognized under copyright law.[5]
The resolution of this issue will likely determine the future legality of AI training practices in India.
B. Applicability of the Fair Dealing Exception
A second critical issue concerns the interpretation of the fair dealing provisions under Section 52 of the Copyright Act.
Indian copyright law recognizes certain exceptions that permit limited use of copyrighted works without authorization. OpenAI’s defense has relied partly on the argument that AI training constitutes a form of research and technological development that should receive protection under fair dealing principles.
ANI disputes this interpretation. It argues that ChatGPT is a commercial product developed and deployed for profit. Consequently, extending the fair dealing exception to commercial AI training could undermine the economic rights of copyright owners and substantially weaken copyright protection.
The case therefore raises an important question: can a doctrine designed for traditional educational and research activities be extended to large-scale commercial AI development?
C. Territorial Jurisdiction
The litigation also presents significant jurisdictional challenges.
OpenAI has argued that its servers and training infrastructure are located outside India. Therefore, according to the company, Indian courts should exercise caution in asserting jurisdiction over activities that allegedly occurred abroad.
ANI maintains that the effects of the alleged infringement are felt within India and involve copyrighted works owned by an Indian entity. As a result, Indian courts possess sufficient territorial nexus to adjudicate the dispute.[6]
The Court’s approach to jurisdiction may influence future cross-border technology disputes involving multinational AI companies.
D. Comparative International Perspective
The ANI litigation is part of a broader global trend.
In the United States, multiple lawsuits have been initiated against AI developers by authors, publishers, artists, and media organizations. One of the most prominent examples is the litigation involving major news publishers and OpenAI concerning the use of copyrighted journalistic content for AI training.
Similarly, European regulators and policymakers have examined whether AI developers should be required to disclose training datasets and obtain licenses for copyrighted materials. The European Union’s evolving AI governance framework reflects increasing concern regarding transparency, accountability, and copyright compliance.
These developments demonstrate that the issues before the Delhi High Court are not unique to India. Rather, they form part of an international effort to reconcile copyright protection with technological innovation. India’s response may therefore contribute to the emerging global discourse on AI governance.
IV. Critical Analysis
The ANI v. OpenAI litigation exposes a fundamental tension between two competing public interests: protecting creators and encouraging innovation.
On one hand, copyright law exists to reward creativity and incentivize the production of original works. Journalists, authors, artists, and publishers invest substantial resources in creating content. If AI companies can freely utilize copyrighted material without authorization or compensation, content creators may suffer economic harm. Such a result could weaken incentives for future creativity and undermine the objectives of copyright law.
The concerns raised by ANI are particularly significant in the context of journalism. News agencies depend on subscriptions, licensing agreements, and syndication arrangements to sustain their operations. If AI systems can generate summaries or reproduce information derived from journalistic content, questions arise regarding market substitution and potential loss of revenue.
On the other hand, excessive restrictions on AI training may impede technological progress. Modern AI systems require enormous quantities of data to function effectively. Requiring developers to negotiate individual licenses for every piece of content used during training may create substantial financial and administrative barriers. Such requirements could disproportionately affect startups and smaller innovators while benefiting only large corporations capable of bearing compliance costs.
The dispute also highlights a broader legislative challenge. The Copyright Act, 1957 was enacted long before the emergence of machine learning, neural networks, and generative AI. Consequently, courts are being asked to apply analog-era legal concepts to digital-age technologies. Although existing principles provide a starting point, they may not adequately address the technical realities of AI training.
A balanced approach is therefore necessary. Absolute immunity for AI developers could undermine the rights of creators, while overly restrictive copyright rules could stifle innovation and technological advancement. The preferable solution may lie in legislative reform rather than judicial improvisation.
India could consider introducing a statutory licensing framework specifically designed for AI training. Such a framework could permit the use of copyrighted materials for training purposes while ensuring fair compensation for rights holders. Similar mechanisms already exist in certain areas of copyright law and may provide a useful model for addressing AI-related concerns.
Additionally, transparency obligations could require AI developers to disclose categories of training data and implement mechanisms through which copyright owners may opt out of future training activities. Such measures would promote accountability while preserving opportunities for innovation.
V. Conclusion
The ANI v. OpenAI litigation represents a landmark moment in the evolution of Indian copyright law. For the first time, an Indian court is being asked to determine whether the use of copyrighted material for training generative AI systems constitutes infringement under the Copyright Act, 1957.
The case raises complex questions concerning reproduction, fair dealing, jurisdiction, and the future relationship between copyright law and artificial intelligence. Its significance extends far beyond the interests of the parties involved. The outcome may influence the development of India’s AI ecosystem, shape the rights of content creators, and contribute to the global debate regarding AI governance.
While the courts will play an important role in resolving the immediate dispute, long-term certainty may ultimately require legislative intervention. Existing copyright frameworks were not designed with generative AI in mind, and the challenges presented by modern machine learning technologies may demand tailored regulatory solutions.
As India seeks to position itself as both a leader in technological innovation and a protector of intellectual property rights, the ANI v. OpenAI case will likely serve as a defining precedent in determining how these objectives can be balanced in the age of artificial intelligence.
VI. References
[1] ANI v OpenAI proceedings before the Delhi High Court concerning the alleged use of ANI’s copyrighted news content for training generative AI models.
[2] Copyright Act 1957, ss 14 and 51.
[3] Copyright Act 1957, s 52.
[4] Reports concerning publisher and media industry participation in the ANI v OpenAI litigation.
[5] R.G. Anand v Delux Films (1978) 4 SCC 118.
[6] Eastern Book Company v D.B. Modak (2008) 1 SCC 1
