Major technology companies, including OpenAI, Google, Microsoft and other global tech giants, are pushing back against India’s proposed new AI copyright framework that would require AI developers to pay royalties for using copyrighted material to train artificial intelligence models. The draft proposal, part of a broader effort to update the country’s copyright laws for the age of generative AI, has sparked industry concern over potential costs, compliance burdens and innovation risks.
What India’s New AI Copyright Rules Propose
India’s government — through the Department for Promotion of Industry and Internal Trade (DPIIT) and other bodies — has drafted a framework aimed at regulating how AI companies train their models using copyrighted content:
- 🔹 Mandatory royalty payments for using copyrighted works in AI training.
- 🔹 A centralised royalty collection system to distribute payments to creators.
- 🔹 Royalties may be applied retroactively to AI models already trained on copyrighted content.
- 🔹 The proposal rejects broad “text and data mining” (TDM) exemptions and voluntary licensing in favour of a “blanket licensing” model.
The working paper argues this approach would ensure fair compensation for creators — including authors, artists and news publishers — whose works are extensively used in AI training.
Tech Giants Raise Objections
However, tech firms and industry groups have expressed significant resistance to parts of the proposal:
🔹 Business Software Alliance (BSA) — representing Microsoft, AWS, IBM, Adobe and others — has urged the Indian government to adopt a more flexible text and data mining (TDM) exception instead of a rigid royalty regime. They say TDM exemptions, widely used in other jurisdictions, enable innovation without compromising copyright protections.
🔹 Nasscom and other industry bodies have warned that a mandatory licensing and royalty system could slow innovation, increase costs for developers, and make India a less attractive market for AI development. Critics argue the framework may disadvantage both global and domestic AI startups.
🔹 AI companies argue that the sheer scale and complexity of training datasets make it technically and economically impractical to track and compensate creators individually under India’s proposed model. A blanket royalty approach could set precedents not seen elsewhere, industry sources say. The Indian Express
Wider Global Context
India’s moves come amid a global debate over AI and copyright:
- In the United Kingdom, OpenAI and Google previously opposed proposed AI copyright rules, favouring broader exceptions for data mining instead of stricter protections.
- Similar discussions are underway in the European Union and the United States, where lawmakers are exploring transparency requirements and linkage between AI and intellectual property rights.
India’s approach, especially its push for compulsory royalty payments and centralised distribution, marks a more interventionist stance than many Western models, reflecting a desire to protect creators while nurturing domestic AI growth.
What Comes Next
The Indian government has opened a public consultation period lasting roughly 30 days, allowing stakeholders — including tech companies, rights holders and industry experts — to submit comments on the proposed rules before they move toward formal legislation.
Depending on the feedback and political will, India’s amended copyright law could become a global benchmark for how fast-growing AI ecosystems handle the tricky intersection of intellectual property, innovation and economic growth.
