The proposed AI framework seeks to formalise royalty funds for coaching datasets, making a unified system supposed to safeguard creators’ rights whereas lowering compliance burdens for builders utilizing copyrighted materials.
India has proposed one of many world’s most far-reaching frameworks for regulating how synthetic intelligence firms practice their fashions on copyrighted content material. The Division for Promotion of Trade and Inner Commerce (DPIIT) on Tuesday launched a draft system that might mandate royalty funds from AI builders in trade for automated entry to all copyrighted works. The royalties could be collected via a central physique representing rights holders and redistributed to creators.
The proposal argues {that a} “necessary blanket licence” would each decrease compliance prices for AI corporations and be sure that authors, musicians, artists and different creators are compensated when their work is used to coach business fashions. Indian officers say the strategy might stop years of litigation and provide readability in a sector the place world courts are nonetheless debating the legality of utilizing copyrighted materials for AI coaching.
The eight-member committee behind the draft, constituted in April, mentioned the system would act as a single-window mechanism, permitting AI firms to keep away from particular person negotiations whereas guaranteeing honest compensation for registered and unregistered creators alike.
World authorized considerations kind backdrop to Indian proposal
India’s transfer comes at a time when firms resembling OpenAI, Google and different AI builders are dealing with lawsuits worldwide over their coaching practices. In India, information company ANI has sued OpenAI within the Delhi Excessive Courtroom, arguing its content material was used with out consent — a case that might decide whether or not AI coaching qualifies as copy or falls beneath “honest dealing.” Related circumstances within the U.S. and Europe have intensified calls for for regulatory readability.
The draft additionally highlights India’s rising position within the world AI market. Citing OpenAI CEO Sam Altman’s remark that India is the corporate’s second-largest market, the committee argues that creators ought to obtain a share of the worth generated by AI corporations working within the nation.
Trade teams warn of innovation dangers
The proposal has triggered resistance from main know-how our bodies. Nasscom submitted a proper dissent, urging India to undertake a broad text-and-data-mining (TDM) exception that might permit coaching on lawfully accessed materials with out necessary licensing. It warned that obligatory funds might gradual innovation and argued for an opt-out mannequin for rights holders.
The Enterprise Software program Alliance echoed these considerations, arguing that limiting AI fashions to licensed datasets might scale back mannequin high quality and amplify biases. It mentioned a transparent TDM exception would higher stability innovation and creator rights.
The committee rejected each opt-out and broad TDM exceptions, calling them both unenforceable or insufficiently protecting. The draft is now open for 30 days of public session earlier than the federal government finalises its regulatory strategy.
Leave a Reply