An expert committee in India has recommended that artificial intelligence (AI) developers must pay royalties for using copyrighted content to train their AI systems. The committee, led by the Department for Promotion of Industry and Internal Trade (DPIIT), released a working paper on December 8, 2025. This proposal introduces a "hybrid model" that would require a mandatory blanket license for AI firms to access lawfully obtained copyrighted works.^[1][2][3][4] The new framework also includes a statutory right for copyright holders to receive payment. This means creators, like musicians, authors, and news publishers, would get compensated for their content used in AI training. The government panel’s report, titled 'One Nation, One License, One Payment: Balancing AI Innovation and Copyright', aims to strike a balance between fostering AI growth and protecting creators' rights. ^[1][5][6][7]### New Royalty System Proposed Under the proposed system, AI developers would gain automatic access to copyrighted content for training purposes, but only after paying mandatory royalties. These royalty payments would begin once the AI tools are commercialized and start generating revenue. A government-appointed committee would set the specific royalty rates. ^[8][9][10]The committee rejected the idea of voluntary licensing, where AI developers negotiate individual deals with content owners. They argued this approach would lead to high transaction costs and unequal bargaining power, especially disadvantaging small creators. The panel also dismissed a blanket "Text and Data Mining" (TDM) exception, which some in the tech industry advocated for, saying it would undermine copyright protections. ^[1][8][3][6]Instead, the panel suggested a centralized, non-profit entity called the 'Copyright Royalties Collective for AI Training (CRCAT)'. This body would be set up by rights holders and officially recognized by the Central government under the Copyright Act, 1957. The CRCAT would be responsible for collecting royalty payments from AI developers and then distributing them fairly to copyright owners. ^{[1][2][8][3][9]}### Global Implications and Industry Concerns If implemented, India would become the only country to establish a statutory licensing regime for AI developers, with royalty rates determined by a government committee. This approach differs significantly from models in the United States, where AI companies often argue that using publicly available data for training falls under "fair use" and does not require payment. It also contrasts with the European Union's "opt-out" model, which the Indian panel found ineffective for creators. ^{[1][11][12][7]}The proposal suggests that copyright owners would not have the option to withhold their works from being used for AI training under this mandatory licensing framework. The committee's report also recommends that royalty payments should apply retroactively. This means AI developers who have already trained their systems on copyrighted content and are earning revenue from them would be obligated to pay the prescribed royalties. ^[1][2][3][7]### Balancing Innovation and Creator Rights The DPIIT-led committee, chaired by Additional Secretary Himani Pande, emphasized that access to large volumes of high-quality data is crucial for AI development. However, it also stressed the importance of ensuring fair compensation for human creators to prevent the weakening of the creative ecosystem. The committee believes this "hybrid model" provides a single window for AI developers to access copyrighted works, simplifies licensing, and reduces transaction costs. ^{[2][8][3][6][10][4]}The recommendations have generated mixed reactions within the industry. Nasscom, a prominent tech industry body representing companies like Google and Microsoft, has formally disagreed with the proposal. Nasscom described the mandatory fee as a "tax or levy on innovation," arguing it could hinder AI development in India. C^[11][12]onversely, content industry representatives have largely supported measures to ensure creators are compensated. ^[3]The working paper has been released for public and stakeholder consultation. Interested parties have 30 days from December 8 to submit their feedback to the DPIIT. The government will then review these comments before deciding on the next steps for this proposed regulatory framework. ^[12][3][4]

AI Insights Hub