India is considering a new royalty system where AI companies, including OpenAI and Google, would compensate content creators for using copyrighted material to train their AI models. The proposal suggests a centralised royalty system, a shift from the 'fair use' approach prevalent in the United States. A government panel, formed in April, suggests AI firms should pay royalties to a central body representing copyright holders, while still having access to Indian content for training.
This initiative aims to resolve copyright conflicts arising from AI models using large-scale data. Unlike the US, where training on publicly available content is considered permissible, India's framework resembles a regulated licensing environment. The Indian panel dismissed the EU's opt-out system, deeming it ineffective, and instead proposes a system where creators can claim revenue from a collective fund if their content is used. Industry faces a 30-day challenge period before government officials review the plan.
Nasscom, a tech industry body, has expressed concerns, viewing the mandatory fee as a 'tax on innovation'. The Motion Picture Association has advocated for maintaining the current copyright law and focusing on licensing. The proposed framework includes a 'one nation, one licence, one payment' model, granting LLM developers access to copyrighted content while ensuring fair compensation for creators.
Related Articles

Google TPU challenges Nvidia
Read more about Google TPU challenges Nvidia →
OpenAI Announces 'Code Red' Focus
Read more about OpenAI Announces 'Code Red' Focus →
OpenAI's Dominance Under Threat
Read more about OpenAI's Dominance Under Threat →
AI Shapes Market Strategies
Read more about AI Shapes Market Strategies →
