India Proposes Mandatory Royalties for AI Training, Targeting OpenAI and Google Operations

India AI Regulation Introduces Mandatory Payment Framework

India has unveiled one of the world's most interventionist approaches to AI copyright, proposing a mandatory royalty system for AI companies that train their models on copyrighted content1

1

. The Department for Promotion of Industry and Internal Trade (DPIIT) released a 125-page working paper on December 8 outlining a framework that would require OpenAI, Google, and other AI companies to pay for content used in training their models3

3

. The proposal comes as governments worldwide develop regulations to resolve AI-related copyright disputes, with AI firms arguing they are making fair use of publicly available material4

4

.

The eight-member committee, formed in late April and chaired by Himani Pande, an additional secretary at DPIIT, argues this blanket licensing system for AI would lower compliance costs while ensuring content creators receive compensation when their work is scraped to train commercial models1

1

. This marks a sharp divergence from jurisdictions like the United States, where AI giants say training models on publicly available data constitutes fair use for which they should not be charged4

4

.

Central Body for Royalty Payments to Streamline Compensation

The proposal establishes a central body to collect and distribute revenue for using work to train AI models. The committee suggests naming this organization the Copyright Royalties Collective for AI Training (CRCAT), which would function as a nonprofit organized by associations of rightsholders2

2

. The working paper also recommends establishing a "Works Database for AI training royalties" that would invite content creators to register their works to receive royalties from CRCAT2

2

.

This single-window system aims to eliminate the need for individual negotiations and enable royalties to flow to both registered and unregistered creators1

1

. The committee argues that a blanket license "aims to provide an easy access to content for AI developers... reduce transaction costs... [and] ensure fair compensation for rightsholders," calling it the least burdensome way to manage large-scale AI training1

1

. Crucially, AI companies would only be required to pay creators for training AI models once they start producing revenue2

2

.

Tech Industry Pushback Against Mandatory Royalty System for AI

Nasscom, the influential tech industry body representing Google and Microsoft, has formally dissented from the proposal, arguing that the mandatory fee amounts to a "tax or levy on innovation". The organization filed a formal dissent arguing that India should instead adopt a broad text and data mining exception that would allow AI developers to train on copyrighted content as long as the material is lawfully accessed1

1

.

The Business Software Alliance, representing global tech firms including Adobe, Amazon Web Services, and Microsoft, pressed the Indian government to avoid a purely licensing regime. It urged India to introduce an explicit text-and-data-mining exception, warning that relying solely on direct or statutory licensing for AI training data may be impractical1

1

. The Motion Picture Association, which represents Netflix and Paramount, previously told the panel there should be no change in copyright law, with efforts focused on licensing instead4

4

.

Why India Matters for Generative AI Companies

The committee points to India's growing importance as a market for generative AI tools to justify its interventionist approach. Citing OpenAI CEO Sam Altman's remark that India is the company's second-largest market after the U.S. and "may well become our largest," it argues that because AI firms derive significant revenue from Indian users while relying on Indian creators' work to train their models, a portion of that value should flow back to those creators1

1

.

OpenAI is currently locked in a court battle triggered by accusations from Indian news agency ANI over use of copyrighted content4

4

. The company has repeatedly denied wrongdoing and said its use of online content amounted to fair use. The Delhi High Court is examining whether AI training is itself an act of reproduction or protected by "fair dealing"1

1

.

Rejecting the Opt-Out Model for Creator Remuneration

The committee firmly rejected the opt-out model advocated by some tech firms, calling it ineffective. While the European Union has stricter rules that allow content owners to opt out of such use, the Indian panel said this approach unfairly forced creators to track down their own work in massive AI datasets4

4

. Instead, if their work is used by AI platforms, they can claim funds from the centralized royalty pool4

4

.

The committee sides firmly with creator remuneration and rejects the push by tech firms for unrestricted text and data mining3

3

. Industry and the public have 30 days to challenge the Indian plan, which faces review by government officials4

4

. AI companies like OpenAI and Google may face new costs as they would be required to pay royalties into a central body for using local content to regulate AI training on copyrighted content5

5

.