India protects data through a blend of copyright, contract, equity, and evolving privacy regulation, without a sui generis database right or codified trade secrets statute. The Copyright Act, 1957 defines tables, compilations, and computer databases as 'literary works' under Section 2(o), and in Eastern Book Company v. D.B. Modak the Supreme Court rejected 'sweat of the brow' in favor of a 'skill and judgment' test. Courts have protected customer lists as original compilations, as in Burlington Home Shopping v. Rajnish Chibber, but raw data itself remains outside copyright.
Trade secret protection is derived from the Indian Contract Act, 1872, equitable principles, breach of confidence doctrine, and Section 72 of the Information Technology Act, 2000, rather than dedicated legislation. Courts have recognized pricing data, marketing strategies, and customer lists as protectable when reasonable confidentiality measures are taken, and consistently enforce NDAs and confidentiality clauses — though the absence of a statute introduces variability and uncertainty in enforcement, an increasingly important gap for companies developing proprietary AI models and training pipelines.
The Digital Personal Data Protection Act, 2023 — promulgated in August 2023 but notified in November 2025 and phasing in to May 2027 — will introduce a consent-based framework for personal data with extraterritorial reach, grounded in the constitutional right to privacy recognized in Justice K.S. Puttaswamy v. Union of India. India has no unified AI law, but in 2024 the Ministry of Electronics and Information Technology issued advisories on generative AI — an initial prior-approval requirement was withdrawn on 15 March 2024 — mandating labeling of AI-generated content.
Key takeaways
- Databases qualify as 'literary works' under Indian copyright law, but protection requires skill and judgment in selection or arrangement, not mere labor.
- With no trade secrets statute, well-drafted NDAs and confidentiality clauses under the Contract Act are the primary safeguard for proprietary data.
- The DPDP Act, 2023 is being operationalized in phases — the DPDP Rules 2025 were notified in November 2025, with remaining obligations binding from May 2027 — imposing a consent-based regime with extraterritorial application to entities serving Indian users.
- MeitY's 2024 advisories signal growing administrative oversight of generative AI — the March 2024 prior-approval requirement was withdrawn within a fortnight, but content-labeling expectations survived and became binding under the IT Amendment Rules 2026.
- Competition authorities are scrutinizing data practices such as exclusionary access to user data as potential abuse of dominance.
Primary legal instruments
AI-assisted summary of the report, editorially reviewed before publication · AI transparency
Protection mechanisms
Report-derived analysis per mechanism. The report text is fixed; committee members with contributor access can revise the rating (with an audit trail) and add notes below each block.
Recent Developments
Live updates contributed by committee members and the AI research bot — separate from the fixed report content above.
Legal consideration of AI copyright implications anticipated
India is poised for a landmark legal decision concerning the copyright implications of generative AI. This development is expected to address the complex interplay between personal data, copyright, and trade secrets in the context of AI training and output. The outcome will significantly shape the legal landscape for AI innovation and data protection in the country.
Delhi High Court clarifies copyright implications for AI training data
The Delhi High Court has issued a significant ruling regarding the use of copyrighted material for training large language models (LLMs). The court determined that OpenAI's storage of news content from Asian News International (ANI) for training ChatGPT did not constitute copyright infringement. This decision marks India's first major judicial pronouncement on the legality of using copyrighted data in AI development, providing crucial clarity for the evolving landscape of AI and intellectual property.
Delhi High Court reserves judgment in ANI v OpenAI, India's first AI training-data copyright case
After 32 hearings ending in late March 2026, the Delhi High Court reserved judgment in Asian News International's copyright suit against OpenAI over the use of ANI's news content to train ChatGPT. The court is weighing whether storing and tokenizing copyrighted works for model training infringes copyright, whether outputs infringe, whether Section 52 fair dealing applies to commercial AI training, and whether Indian courts have jurisdiction over US-hosted servers. The two court-appointed amici reportedly diverge on the core technical question, and the forthcoming verdict will set India's first judicial rule on text-and-data mining for AI.
DPIIT closes extended consultation on Generative AI & Copyright Working Paper proposing statutory licensing of training data
DPIIT extended by 30 days, to 6 February 2026, the public consultation on its committee's Working Paper on Generative AI and Copyright (Part I). The paper proposes a hybrid model under which AI developers receive blanket statutory licences to train on copyrighted datasets, with royalty obligations triggered at commercialization and administered through a proposed Copyright Royalties Collective for AI Training, with government-set rates subject to judicial review. Developers would also file training-data disclosure forms identifying the datasets used.