<\/p>\n
India’s government is reviewing its Copyright Act 1957 in response to generative artificial intelligence’s impact on creative industries, amid legal challenges and calls for clearer regulation of AI training practices.<\/p>\n<\/div>\n
The rapid ascent of generative artificial intelligence has forced a reappraisal of copyright law in India, with the central government now explicitly examining whether the Copyright Act, 1957 remains fit for purpose in an era when machines ingest and repurpose vast stores of human expression. According to a written reply in Parliament by Union Minister of State for Commerce and Industry Jitin Prasada, the Department for Promotion of Industry and Internal Trade (DPIIT) has convened an eight\u2011member expert committee, established on April 28, 2025, and has published Part 1 of a working paper on the use of copyrighted content in AI training, opening it for stakeholder feedback. [1]<\/a><\/sup>[2]<\/a><\/sup><\/p>\n The committee\u2019s remit, as set out by the government, includes analysing legal and policy challenges arising from AI\u2019s use within the copyright framework, assessing whether existing statutory provisions are adequate, and recommending changes where necessary. The government has signalled that these findings will feed into a forthcoming white paper that could guide legislative or administrative responses. [1]<\/a><\/sup>[2]<\/a><\/sup><\/p>\n At the heart of the debate is the training of AI models on copyrighted material scraped from the internet and other sources. Industry submissions to the government have urged clarification of the legality of such practices, including possible amendments to the Copyright Act and clearer definitions of \u201cfair dealing,\u201d while calling for public consultation on governance issues including transparency and deepfakes. According to the industry, legal certainty is essential to promote adoption and investment in AI. [3]<\/a><\/sup><\/p>\n Creators and rights\u2011holders say the current state of play amounts to uncompensated appropriation of creative labour. Artists, journalists, photographers, publishers and their representative organisations have argued that generative AI companies are free\u2011riding on decades of human creativity by using copyrighted works as raw material without licensing or remuneration; this backlash is manifesting in litigation and collective bargaining efforts worldwide. The Indian context is already seeing related legal action, most notably a copyright suit by ANI Media against OpenAI alleging unauthorised use of news content to train ChatGPT. [1]<\/a><\/sup>[3]<\/a><\/sup>[6]<\/a><\/sup><\/p>\n That litigation has also exposed jurisdictional and cross\u2011border complications. OpenAI has responded in Indian proceedings by arguing that its model training and data storage occurred outside India and therefore fall beyond the territorial reach of the Indian Copyright Act; it has further said it cannot delete training data where it is bound by foreign legal or operational constraints. These positions raise acute questions for enforcement if training datasets are dispersed and operations are global. [4]<\/a><\/sup>[5]<\/a><\/sup>[6]<\/a><\/sup><\/p>\n A second, equally fraught issue is whether outputs generated by AI can attract copyright protection at all. Traditional Indian doctrine links originality to human skill, labour and judgment; if an algorithm merely recombines existing works, courts and policymakers will have to decide whether such outputs qualify as \u201coriginal\u201d or belong to the public domain. Conversely, denying protection could chill commercial deployment by removing exclusivity for businesses that invest in machine\u2011generated content. The hybrid reality, where human prompting, selection and editing intermix with automated generation, complicates any bright\u2011line test. [1]<\/a><\/sup><\/p>\n The problem stretches beyond text to images, music and video. Image\u2011generation models that produce work \u201cin the style of\u201d living or deceased artists, and music systems that emulate genres or performers, test the boundary between lawful influence and actionable imitation. In India\u2019s burgeoning NFT and digital\u2011art ecosystems these questions have immediate market consequences; unclear ownership rules risk disputes over authenticity, monetisation and the distribution of value between human creators and platform or model owners. [1]<\/a><\/sup><\/p>\n Policy options under consideration range from calibrated exceptions for model training, to collective licensing mechanisms, to transparency and provenance obligations for AI developers, and potentially to novel rights tailored for AI\u2011generated works. Any framework will need to balance protection for authors, journalists and artists against regulatory burdens that could concentrate AI capabilities in a handful of deep\u2011pocketed firms. DPIIT\u2019s expert committee and the forthcoming white paper will be pivotal in translating these trade\u2011offs into concrete proposals. [1]<\/a><\/sup>[2]<\/a><\/sup>[3]<\/a><\/sup><\/p>\n For now, India\u2019s legal system and market participants are operating in a period of uncertainty that is playing out in courts, regulator consultations and industry submissions. The outcome will shape not only how value is allocated between humans and machines in India\u2019s creative industries, but may also influence international approaches to AI and copyright governance in a globally interconnected digital economy. [1]<\/a><\/sup>[3]<\/a><\/sup>[4]<\/a><\/sup>[6]<\/a><\/sup><\/p>\n ##Reference Map:<\/p>\n\n