India proposes charging OpenAI, Google for training AI on copyrighted content

Core Viewpoint - India's proposal for a mandatory royalty system for AI companies training on copyrighted content could significantly alter the operational landscape for firms like OpenAI and Google in a rapidly growing market [1][3]. Group 1: Proposal Details - The Indian government has introduced a framework that allows AI companies to access copyrighted works for training in exchange for royalties paid to a new collecting body, which will distribute payments to creators [2][4]. - This "mandatory blanket license" aims to reduce compliance costs for AI firms while ensuring fair compensation for rights holders, including writers, musicians, and artists [5][6]. - The proposal is positioned as a solution to ongoing legal uncertainties regarding the use of copyrighted material for AI training, contrasting with the more fragmented approaches seen in the U.S. and EU [3][4]. Group 2: Market Context - India's significance as a market for generative AI tools is highlighted, with OpenAI's CEO noting that India is the company's second-largest market and may become the largest [6]. - The proposal is seen as a way to ensure that a portion of the revenue generated by AI firms from Indian users is returned to local creators [6]. Group 3: Industry Reactions - The proposal has faced dissent from industry bodies like Nasscom, which advocates for a text-and-data-mining exception instead of a mandatory licensing regime, arguing it could hinder innovation [12]. - The Business Software Alliance has also expressed concerns, suggesting that a licensing-only approach may not yield optimal outcomes and could limit the quality of AI models [13]. - The committee has dismissed the idea of a broad text-and-data-mining exception, proposing instead a hybrid model that grants automatic access to copyrighted works while requiring royalty payments [14]. Group 4: Next Steps - The Indian government has opened the proposal for public consultation, allowing stakeholders 30 days to provide feedback before finalizing recommendations [15].