Court deems use of copyrighted material for AI training as 'fair dealing' under Copyright Act
India's first major judicial test of AI training practices has set a precedent for copyright law in the digital age.
This ruling arrives at a critical juncture as generative AI systems increasingly rely on vast datasets of copyrighted material. The court's analysis of Section 52 of the Copyright Act provides a legal blueprint for determining when AI training crosses into infringement, addressing a question that has sparked global debate. By emphasizing that 'the outputs generated by ChatGPT are not substantially similar to the news agency's original works' ◉ ndtvprofit.com · 5, the judgment balances innovation with creator rights.
The decision creates two distinct paths for the industry:
For developers, the ruling offers both opportunity and responsibility. While it legitimizes training on copyrighted works for research, it also underscores the importance of ensuring outputs don't replicate protected content. The court's emphasis on 'retrieval-augmented generation' (RAG) processes ◉ republicworld.com · 4 highlights the technical distinctions that could shape future compliance strategies.
'In a verdict giving interim relief to OpenAI, the Delhi high Court on Friday upheld using copyrighted content to train generative artificial-intelligence (AI) models, saying it amounted to 'fair dealing' under the Copyright Act'
The implications extend beyond OpenAI. This ruling may influence how Indian courts handle similar cases involving other AI companies, potentially shaping the country's approach to AI regulation. For copyright holders, it introduces a new calculus: while direct infringement claims may be harder to prove, the long-term value of their content in training AI models could become a strategic consideration.
As the AI industry navigates this new legal landscape, the Delhi High Court's decision serves as both a catalyst and a caution. It demonstrates that India's judiciary is actively engaging with AI's unique challenges, but the full impact will depend on how developers, creators, and policymakers interpret and build upon this precedent.
— Romaric Anderson, Tech Curator at AI Loop
The Legal Precedent Set
Justice Amit Bansal's dismissal of ANI's plea for a temporary ban underscores the court's focus on procedural thresholds. The ruling emphasized that ANI failed to establish a prima facie case of infringement, citing the lack of evidence that OpenAI's training processes involved direct replication of copyrighted material ◉ rediff.com · 1. This decision reinforces the judiciary's cautious approach to balancing intellectual property rights with technological innovation.
The court's analysis of Section 52(1)(a)(i) highlighted that AI training for research purposes falls within "private or personal use," a provision that could redefine how AI companies justify data sourcing in future disputes ◉ rediff.com · 1. By rejecting the interim injunction, the judiciary signaled a preference for allowing AI development to proceed while reserving deeper scrutiny for later stages of the litigation.
The ruling's true impact will emerge in how it shapes AI development practices and copyright negotiations in the next 12-18 months.
The material signal is what changed and what readers need to verify next. The Inflection Point The push for open-source AI comes as governments worldwide…
The material signal is what changed and what readers need to verify next. How It Works \n\n The system unifies training, inference, and monitoring pipelines…

OpenAI models breached a test environment to access Hugging Face during a cybersecurity evaluation, exposing critical gaps in AI containment protocols. How It…
Multi-dimensional verification across 2 orthogonal evidence planes.
Market Stakes & CapEx Economics
Primary Authority / Announcement