DPIIT Releases Working Paper on Copyright–AI Policy, Proposes Hybrid Licensing

The working paper marks a significant step in India’s formulation of a structured and balanced AI–copyright policy that protects creators while enabling innovation.

The Department for Promotion of Industry and Internal Trade (DPIIT) has released Part 1 of its working paper examining the rapidly evolving relationship between generative artificial intelligence (AI) and copyright law. The paper incorporates the findings and recommendations of an eight-member expert committee established on April 28, 2025, to evaluate whether India’s current legal framework adequately addresses the challenges posed by generative AI technologies.

The working paper marks a significant step in India’s formulation of a structured and balanced AI–copyright policy that protects creators while enabling innovation.

---

Background and Purpose of the Committee

With AI systems increasingly relying on massive datasets—including copyrighted material—the Committee was tasked with assessing:

• Whether existing copyright laws sufficiently regulate AI training and output.

• Global approaches adopted by other jurisdictions.

• Potential policy frameworks balancing creators’ rights and technological advancement.

• Amendments required, if any, to strengthen India’s copyright ecosystem.

The working paper reflects an extensive analysis of legal, economic and technological considerations shaping the AI landscape.

---

Assessment of Existing Global Approaches

The Committee examined several international models, including:

1. Blanket Exemptions for AI Training

Where all copyrighted content can be freely used for training AI. Concern: Risks severely undermining creators’ rights and economic incentives.

2. Text and Data Mining (TDM) Exceptions

Permitting data use with or without an opt-out mechanism. Concern: Opt-out systems may be ineffective; opt-in may be burdensome.

3. Voluntary Licensing Systems

Where AI developers negotiate individual licenses. Concern: Impractical at scale due to transaction complexity.

4. Extended Collective Licensing

Collective bodies license content on behalf of creators. Concern: Suitability in the Indian context remains uncertain.

All existing models were found to present limitations—either by weakening creator protection or by imposing unreasonable compliance burdens on innovators.

---

Rejection of Zero-Price Licensing

The Committee firmly rejected a zero-price blanket licensing regime, arguing that:

• It would erode long-term incentives for human creators.

• It risks reducing the production of high-quality human-generated content.

• It creates an uneven playing field between tech companies and individual creators.

The working paper stresses that any future policy must safeguard economic and moral rights of creators while ensuring that AI development is not stifled.

---

Proposed Hybrid Licensing Framework

To address the shortcomings of existing models, the Committee proposes an innovative hybrid approach, aimed at ensuring legal clarity, fairness, and scalability.

1. Blanket Licence for Training Data Usage

• AI developers may use all lawfully accessed content for training purposes.

• No need for individual negotiations or prior approvals.

• Reduces compliance burden and supports innovation.

2. Royalty Payments Triggered Only on Commercialisation

• Royalties apply only when AI tools are commercially deployed, not during research or development.

• Rates would be determined by a government-appointed committee.

• The mechanism allows judicial review, ensuring transparency and fairness.

3. Centralised Royalty Collection and Distribution

• A single national mechanism would handle:

  • Royalty collection

  • Revenue allocation

  • Dispute resolution

• This structure aims to:

  • Reduce transaction costs

  • Protect small and emerging AI developers

  • Ensure creators receive fair compensation

The hybrid model seeks to maintain a balance between fostering innovation and ensuring that content creators are rewarded for their contributions.

---

Acknowledgements

The paper recognizes the significant contributions of Dr. Raghavender G R, whose expertise played a key role in preparing the document. Additional support was provided by Ms. D. Sripriya, Mr. Kushal Wadhawan, and Ms. Priyanka Arora, who assisted the Committee in drafting and research.

---

Public Consultation Open for 30 Days

DPIIT has opened the working paper for public and stakeholder consultation for 30 days, inviting inputs on the proposed hybrid model. The full document is available at: https://www.dpiit.gov.in/static/uploads/2025/12/ff266bbeed10c48e3479c941484f3525.pdf

Feedback received will help shape India’s long-term policy approach to AI governance and copyright protection.