India's Department for Promotion of Industry and Internal Trade (DPIIT) has proposed a mandatory royalty framework for artificial intelligence (AI) companies training models on copyrighted content. The move could significantly affect major AI developers, including OpenAI and Google, within what has become one of their key and fastest-growing global markets.
Released on Tuesday, the proposed "mandatory blanket license" would grant AI companies access to copyrighted works for training. In return, these firms would pay royalties into a new collecting body, composed of rights-holding organizations, which would then distribute payments to creators. The DPIIT framework suggests this approach aims to reduce compliance costs for AI developers while ensuring creators are compensated when their work is utilized for commercial AI model training.
This proposal emerges amid increasing international scrutiny over AI companies' use of copyrighted material, a practice that has led to lawsuits from various rights holders in the United States and Europe. While courts and regulators in those regions debate "fair use" definitions, India's framework represents an interventionist approach by mandating access and payment. An eight-member government committee, formed in April, stated the system is designed to bypass prolonged legal uncertainty and guarantee early compensation for creators, citing India's growing importance as an AI market.
However, the proposal has encountered pushback from industry groups. Nasscom, an Indian technology industry body representing firms including Google and Microsoft, filed a formal dissent. Nasscom advocated for a broad text-and-data-mining (TDM) exception, arguing it would allow AI developers to train on lawfully accessed copyrighted content without mandatory licensing. The organization warned that a mandatory licensing regime could impede innovation and suggested rights holders should have an opt-out mechanism. Similarly, the Business Software Alliance (BSA), representing global tech firms like Adobe and Amazon Web Services, urged an explicit TDM exception, citing concerns that a purely licensing-based model could reduce AI model quality by limiting training data sets.
The government committee, in response to these concerns, stated that broad TDM exceptions or opt-out models compromise copyright protections or are challenging to enforce. It instead proposed a "hybrid model" combining automatic access to lawfully available copyrighted works with mandatory royalty payments through the central collecting body. The Indian government has opened a 30-day public consultation period for stakeholders to submit feedback before final recommendations are formulated. OpenAI and Google have not publicly commented on the proposal.
