Reference Article: Editorial | The Hindu – Cyber crackdown: On the investigation into cyber-crime
UPSC Relevance:
– GS III – Science and Technology
Large Language Models (LLMs) have advanced rapidly due to two key drivers: improved machine learning techniques and vast access to online text, data, and multimedia for training. AI developers claim that publicly available Internet content should be free to use for training, even though non-AI reproduction requires consent or licensing. This has sparked a global clash between AI companies and content creators, including news outlets, publishers, and the entertainment industry.
India’s Proposal: Mandatory Licensing Framework
- DPIIT’s working paper proposes a copyright society-like mechanism:
- AI scraping tools may freely scrape public Internet content.
- A non-profit licensing body would collect royalties from AI developers.
- Royalties would be based on revenues earned by AI models trained on Indian content.
- Rationale:
- Opt-out regimes are impractical for most content creators.
- AI models typically do not reproduce content verbatim but synthesise new outputs.
- A national-level solution prevents India’s AI ecosystem from being handicapped by heavy permissions or litigation.
Challenges and Open Questions
- Royalty distribution concerns:
- Small publishers may receive the same remuneration per unit of content as large media houses despite unequal investment.
- Legal ambiguity persists:
- Global lawsuits between AI firms and publishers offer no consistent judicial precedent.
- The committee rightly notes that waiting for litigation outcomes is risky and delays needed regulation.
Significance of the Proposal
- Offers a balanced approach: ensures remuneration without throttling AI innovation.
- Helps India avoid a regulatory vacuum that would benefit AI hyperscalers at the expense of creators.
- Encourages a collaborative, negotiated framework, supported by the government.
- Aligns India with emerging global norms where the rights of creators and technological innovation must co-exist.
Conclusion
The DPIIT working paper marks an important step toward resolving a contentious AI–copyright conflict. While the proposed licensing system is imperfect, having a structured mechanism is better than a regulatory void that leaves creators uncompensated and enables unchecked AI expansion. A refined framework, supported by government action and judicial oversight, will be crucial for India to balance innovation, content rights, and fairness in the digital economy.