India has proposed a compulsory royalty system for AI firms that prepare their fashions on copyrighted content material — a transfer that would reshape how OpenAI and Google function in what has already change into one in every of their most vital and fastest-growing markets globally.
On Tuesday, India’s Division for Promotion of Trade and Inner Commerce launched a proposed framework that will give AI firms entry to all copyrighted works for coaching in change for paying royalties to a brand new amassing physique composed of rights-holding organizations, with funds then distributed to creators. The proposal argues that this “obligatory blanket license” would decrease compliance prices for AI corporations whereas making certain that writers, musicians, artists, and different rights holders are compensated when their work is scraped to coach business fashions.
India’s proposal comes amid mounting issues in international markets over how AI firms prepare their fashions on copyrighted materials, a observe that has triggered lawsuits from authors, information organizations, artists, and different rights holders within the U.S. and Europe. Courts and regulators are nonetheless weighing whether or not such coaching qualifies as truthful use, leaving AI corporations working below authorized uncertainty and permitting them to quickly increase their enterprise with out clear rules.
Not like the U.S. and the European Union, the place policymakers are debating transparency obligations and fair-use boundaries, India is proposing probably the most interventionist approaches but by giving AI firms automated entry to copyrighted materials in change for obligatory cost.
The eight-member committee, shaped by the Indian authorities in late April, argues the system would keep away from years of authorized uncertainty whereas making certain creators are compensated from the outset.
Defending the system, the committee says in a 125-page submission (PDF) {that a} blanket license “goals to supply an easy accessibility to content material for AI builders… cut back transaction prices… [and] guarantee truthful compensation for rightsholders,” calling it the least burdensome option to handle large-scale AI coaching. The submission provides that the one amassing physique would perform as a “single window,” eliminating the necessity for particular person negotiations and enabling royalties to move to each registered and unregistered creators.
The committee additionally factors to India’s rising significance as a marketplace for GenAI instruments. Citing OpenAI CEO Sam Altman’s comment that India is the corporate’s second-largest market after the U.S. and “might nicely change into our largest,” it argues that as a result of AI corporations derive vital income from Indian customers whereas counting on Indian creators’ work to coach their fashions, a portion of that worth ought to move again to these creators. That, it says, is a part of the rationale for establishing a “balanced framework” that ensures compensation.
Techcrunch occasion
San Francisco
|
October 13-15, 2026
India’s proposal lands amid intensifying authorized battles worldwide over whether or not AI firms can lawfully use copyrighted materials to coach their fashions.
In India, information company ANI sued OpenAI within the Delhi Excessive Court docket, arguing its articles have been used with out permission — a case that has prompted the court docket to look at whether or not AI coaching is itself an act of replica or protected by “truthful dealing.” Courts within the U.S. and Europe are confronting related disputes, with creators alleging that tech firms have constructed their fashions on unlicensed content material.
AI proposal sees pushback and dissent
Not everyone seems to be satisfied by the Indian authorities’s proposed mannequin, although.
Nasscom, the business physique representing expertise corporations together with Google and Microsoft, filed a proper dissent arguing that India ought to as an alternative undertake a broad text-and-data-mining exception that will permit AI builders to coach on copyrighted content material so long as the fabric is lawfully accessed. It warned {that a} obligatory licensing regime may sluggish innovation and mentioned rightsholders who object must be allowed to decide out somewhat than pressure firms to pay for all coaching information.
The Enterprise Software program Alliance, which represents international tech corporations together with Adobe, Amazon Net Providers, and Microsoft, pressed the Indian authorities to keep away from a purely licensing-based regime. It urged India to introduce an express text-and-data-mining exception, arguing that “relying solely on direct or statutory licensing for AI coaching information could also be impractical and will not yield one of the best outcomes.”
Limiting AI fashions to smaller units of licensed or public-domain materials, BSA warned, may cut back mannequin high quality and “improve the danger that outputs merely replicate developments and biases of the restricted coaching information units,” including {that a} clear TDM exception would higher stability innovation and rights holders’ pursuits.
The committee didn’t contemplate each a broad text-and-data-mining exception and an opt-out mannequin, arguing that such techniques both undermine copyright protections or are unimaginable to implement. As an alternative, it proposed a “hybrid mannequin” that will grant AI corporations automated entry to all lawfully obtainable copyrighted works whereas requiring them to pay royalties into the central amassing physique that distributes the proceeds to creators.
The Indian authorities has now opened the proposal for public session, giving firms and different stakeholders 30 days to submit their feedback. After reviewing the suggestions, the committee will finalize its suggestions earlier than the framework is taken up by the federal government.
OpenAI and Google didn’t reply to requests for feedback.
Leave a Reply