India has proposed a mandatory royalty system for AI companies that train their models on copyrighted content – a move that has become one of its most important and fastest-growing markets.
On Tuesday, the Department of India for the promotion of industry and internal trade released The proposed framework that will give AI access to all creative works to train them to generate payments for bodies that match the collecting organization with hardships for creators. The proposal argues that “this mandatory bliking license will lower costs for AI companies while ensuring that these writers, musicians, artists, and writers, and writers, and writers, and writers, and writers, and writers, and writers, and writers, and writers, and writers, and writers, and writers, and writers, and writers, and writers, and writers, and writers, and writers, and writers, and writers, and writers, and writers, and writers, and writers, and writers, and writers, and writers, and writers, and writers, and writers, and writers, and writers, and writers, and writers, and writers, and writers, and writers, and writers, and writers writers and writers
India’s proposal has been in a format embedded in the global market about AI companies that train models for material that can lead to lawsuits, news organizations, artists, and other Rights rights in the US and Europe. Courts and regulators are still considering whether such training qualifies as fair use, leaving active AI companies in limbo and allowing them to rapidly expand their businesses.
Unlike us and the European Union, where the policy rules contradict the obligation of transparency and best use, India proposes one of the AI approaches to automatically generate mandatory payments.
The eight-member committee, formed by the Indian government in late April, argued the system would avoid uncertainty while ensuring creators were compensated from the start.
Defending the system, the committee said in a 125 page submission (PDF) that the blanket license “aims to provide easy access to content for ai developers … (and) guarantee the same way to manage large AI training. The submission adds that the single collecting body will be a “single window,” eliminating the need for individual negotiations and enabling royalties to flow to registered creators.
The committee also gave value to India very important as a market for genai toolsSee rank-. Managing Notes Openai CEO Sam Altman that India is the company’s largest market after the US and “maybe the biggest,” so it may be the greatest experience to train the Indian model, because there is a part that must be returned to the creators. This, he said, is part of the rationale for establishing a “balanced framework” that ensures compensation.
TechCrunch events
San Francisco
I’m fat
October 13-15, 2026
India’s proposal lands in the middle of a worldwide battle over whether AI companies can use copyrighted material to train models.
In India, Ani News Agency SUED Openai in Delhi High Courtobjected to the article being used without permission – an incident that led the court to examine whether ai training is a reproductive act or protected “fair.” Courts in the US and Europe Facing the same hostilitywith the creator asserting that the tech company owns it build models on unlicensed contentSee rank-.
The AI proposal saw pushback and was not taken up
Not everyone is convinced by the model proposed by the Indian government, though.
Nasscom, the industry body that represents technology companies including Google and Microsoft, filed a formal exemption to exchange II that distributes content legally. Given that the compulsory licensing regime can slow down innovation and say that hakistas who object should be allowed to choose instead of those who have to pay for all the training data.
The Business Software Alliance, which represents global technology companies including Adobe, Amazon Web Services, and Microsoft, is being pressured by the Indian government not to have licensees. It sought India’s exemption from introducing explicit text-and-data, saying that “relying on direct or legal licenses for AI training data may not yield the best results.”
Limiting the AI model to various licensed or public materials, the BSA warned, could reduce the quality of the limited data, “adding a clearer balance exclusion and a clearer balance interest.”
The Committee did not consider the broad text-and-data-mining exception, arguing that the system would undermine copyright protection or be impossible to implement. Instead, it proposes a “hybrid model” that would automatically give AI characters access to all copyrighted works in a legal way while asking for royalties to entertain creators.
The Indian government has now opened the proposal for public consultation, giving companies and other winners 30 days to submit comments. After reviewing the feedback, the committee will finalize its recommendations before the framework is taken up by the government.
Openai and Google did not respond to requests.