<\/p>\n
India is moving towards a comprehensive copyright regime that would require AI developers to pay royalties for training models on copyrighted works, sparking debate over innovation and creator rights.<\/p>\n<\/div>\n
India is moving towards a far-reaching copyright regime that would force artificial\u2011intelligence developers to pay royalties for training models on copyrighted works, under a government\u2011endorsed \u201cOne Nation, One Licence, One Payment\u201d blanket\u2011licence model. The proposal, set out in a working paper from the Department for Promotion of Industry and Internal Trade (DPIIT), would create a centralised mechanism to collect and distribute payments to creators and require AI firms to rely only on lawfully accessed material while filing detailed summaries of training datasets. [1]<\/a><\/sup>[5]<\/a><\/sup>[6]<\/a><\/sup><\/p>\n Under the plan, a new body , provisionally named the Copyright Royalties Collective for AI Training (CRCAT) , would administer the licence, collect fees from developers and allocate royalties to rights\u2011holders across sectors including publishing, music, film, journalism and visual arts. The CRCAT is envisaged as a government\u2011designated, non\u2011profit umbrella organisation formed by copyright societies and collective management organisations. [1]<\/a><\/sup>[4]<\/a><\/sup>[5]<\/a><\/sup>[6]<\/a><\/sup><\/p>\n The working paper favours flat, revenue\u2011linked percentage fees tied to global earnings of commercial AI systems, rather than use\u2011based or per\u2011work micro\u2011payments. Those rates would be set by a government\u2011appointed committee, reviewed roughly every three years and subject to judicial challenge, according to the proposal. DPIIT officials have said the fee calculation would look at global revenue, not solely Indian turnover, broadening the financial impact on multinationals. [1]<\/a><\/sup>[7]<\/a><\/sup>[5]<\/a><\/sup><\/p>\n Crucially, the proposal would apply retroactively to developers that have already trained profitable models on copyright\u2011protected material , a measure framed by policymakers as corrective for the creative ecosystem but one that would raise complex legal and compliance questions for firms that have relied on prevailing fair\u2011use or opt\u2011out approaches elsewhere. [1]<\/a><\/sup>[2]<\/a><\/sup><\/p>\n Proponents within government argue the blanket\u2011licence avoids the transaction costs, bargaining asymmetries and fragmentation of voluntary bilateral deals, ensuring dependable access for developers while guaranteeing statutory remuneration for creators. The working paper explicitly rejects voluntary, piecemeal licensing as insufficient to protect smaller creators and to provide reliable data access for AI developers. [3]<\/a><\/sup>[5]<\/a><\/sup><\/p>\n Industry groups and some rights holders have pushed back. Trade association Nasscom and the Motion Picture Association have warned the model could act as a tax on innovation and stifle investment, arguing that mandatory levies and centralised rate\u2011setting risk chilling effects on AI development. Other stakeholders favour licensing frameworks that preserve negotiation flexibility, and some large publishers have already pursued individual licences with developers. The government has opened a 30\u2011day window for public and industry comments before the proposal proceeds to final review. [2]<\/a><\/sup>[3]<\/a><\/sup><\/p>\n The Indian proposal diverges from international approaches: it contrasts with the United States\u2019 general acceptance of training on publicly available data under \u201cfair use\u201d doctrines and departs from the EU\u2019s more nuanced opt\u2011out and rights\u2011management efforts. Japan has taken an even more permissive position. India\u2019s model reflects a broader assertion of national control over data\u2011related rights and a willingness to impose statutory solutions to balance creative remuneration and technological development. [2]<\/a><\/sup>[6]<\/a><\/sup><\/p>\n Implementing the scheme would pose practical challenges: setting administratively fair rates, auditing compliance, tracking the provenance and use of training data, and allocating distributions equitably across millions of works and diverse creator classes. The plan proposes dataset reporting requirements from developers to aid monitoring, but industry lawyers predict litigation over retroactivity, scope and constitutional issues is likely if the framework becomes law. [1]<\/a><\/sup>[5]<\/a><\/sup>[6]<\/a><\/sup><\/p>\n As the consultation period closes and the government considers submissions, the debate crystallises around competing objectives , ensuring creators are paid for commercial reuse of their work, versus preserving a regulatory environment conducive to innovation and investment in AI. How policymakers reconcile those aims will determine whether India\u2019s model becomes a template for others or a contested outlier in global AI governance. [2]<\/a><\/sup>[3]<\/a><\/sup>[5]<\/a><\/sup><\/p>\n ##Reference Map:<\/p>\n\ud83d\udccc Reference Map:<\/h3>\n
\n