India proposes mandatory AI training licence with central royalty body to balance innovation and copyright

India Introduces Mandatory Framework for AI Training and Copyright

The Department for Promotion of Industry and Internal Trade (DPIIT), under the Ministry of Commerce & Industry, has released a 125-page working paper titled "One Nation One Licence One Payment: Balancing AI Innovation and Copyright" that proposes a mandatory blanket licence for AI developers 2

2

. The framework addresses AI copyright issues by requiring companies like Google, OpenAI, and Anthropic to pay rightsholders for copyrighted content used in training commercial AI models 2

2

.

Under this proposed model, AI developers would gain automatic rights to use any lawfully accessed copyrighted works for training without negotiating individually with copyright owners 4

4

. The eight-member committee convened by DPIIT rejects unfettered access to copyrighted data for AI training without compensation, arguing that allowing free use would erode incentives for content creators including authors, artists, and journalists 2

2

.

CRCAT Emerges as Central Gatekeeper for AI Training Royalties

The proposal introduces the Copyright Royalties Collective for AI Training (CRCAT), a central royalty-collection body that would function as the single gatekeeper enforcing rates, ensuring compliance, and distributing payments to creators 5

5

. CRCAT would operate as a nonprofit designated by the central government under the Copyright Act, handling four core functions: collecting royalties from AI developers, distributing funds to collective management organizations (CMOs) and copyright societies, enforcing compliance, and operating the Works Database that determines payout eligibility 5

5

.

The governing board includes one representative from each member organization and temporary representatives from unorganized sectors until they form their own CMO 5

5

. This centralized mechanism aims to simplify the AI training license process by preventing developers from negotiating with thousands of individual creators, though it concentrates operational power in a single organization that does not yet exist 5

5

.

Revenue-Based Royalties and Retroactive Payments

The framework proposes that remuneration for copyright holders be tied to commercialization rather than training itself. Royalties will be paid only when AI models using the data for training are commercialized, calculated as a flat percentage of global revenue earned from the AI system 2

2

5

5

. A government-appointed Rate Setting Committee, not CRCAT, would decide royalty rates and review them every three years 5

5

. Significantly, the paper proposes that royalties should be retroactive, meaning firms that have already used Indian copyrighted works to train models for commercial deployment would also be liable for payment under the new regime 2

2

. The draft rejects alternative approaches such as a "zero-price licence" or an "opt-out" text-and-data-mining exception, with committee members arguing both create unfair burdens on creators, especially those from smaller organizations without resources to police AI datasets 2

2

.

Tech Industry Raises Feasibility Concerns

The strongest early pushback is coming from India's technology industry, which argues that the framework misunderstands the mechanics of AI training and risks burdening an emerging AI ecosystem still taking shape 1

1

.

While the Ministry of Electronics and Information Technology has endorsed DPIIT's proposed framework, concerns about fair use, innovation, and legal framework adequacy persist 1

1

. Under the proposed model, while creators lose the ability to opt out of AI training altogether, they receive a statutory right to remuneration through the new central body 4

4

. The framework requires AI developers to file a Training Data Disclosure Form summarizing broad categories of training data used, though they need not list specific works, datasets, or URLs 5

5

. Creators must register their works in a sector-specific Works Database, and only registered works receive payment, even though unregistered works can still be used in training 5

5

.

Next Steps and Future Implications

DPIIT has opened the draft for stakeholder consultation, inviting feedback over the next 30 days 2

2

. The committee is expected to release its second working paper on copyrightability of AI generated content and authorship in about two months 3

3

. The eight-member panel headed by additional secretary Himani Pande was tasked to identify copyright issues raised by AI systems, examine the existing legal framework, assess its adequacy, and recommend changes if necessary 3

3

. The central challenge lies in how to protect intellectual property in underlying human-created works without stifling technological advancement 3

3

. The proposal's success depends heavily on whether collective management organizations emerge across all creative sectors and whether the framework can balance rightsholders' interests with the operational realities of AI development.