A recent government proposal to resolve copyright disputes around artificial intelligence (AI) training data has sparked significant debate. Critics argue the plan could hinder India's potential to become a global AI leader by imposing heavy costs and relying on flawed systems.
The Core of the Controversy
On December 8, 2025, the Generative Artificial Intelligence and Copyright Committee, operating under the Department for Promotion of Industry and Internal Trade (DPIIT), released a consultation paper. The committee, led by Meghna Bal of the Esya Centre, aimed to address the fierce global debate between AI developers and copyright holders.
Rights holders, including news publishers, authors, and comedians, contend that using their copyrighted works to train AI models without permission or payment is illegal. AI companies counter that training involves learning patterns and statistical relationships, not reproducing protected works, and qualifies as transformative use—a legal defense in India.
The Proposed Blanket Licensing Regime
The committee's solution is a blanket mandatory license for all copyrighted works used in AI training. Under this scheme, every AI model trained on such material would pay a flat rate based on a percentage of its global revenue. The rule would apply retroactively, forcing existing, revenue-generating models to start paying.
A government-appointed committee would set the rates, but the collection and distribution of fees would be handled by a new body: the Copyright Royalties Collective for AI Training, a consortium of existing Collective Management Organisations (CMOs).
Four Critical Flaws in the Plan
While marketed as a win-win, the proposal faces sharp criticism on multiple fronts.
First, it centralises power in CMOs, institutions with a troubled history in India's copyright landscape. Past issues include extorting users, lacking transparency, and colluding to withhold royalties from artists. Despite 2012 amendments to the Copyright Act to curb abuses, problems persisted. Entrusting these entities with a vast, mandatory AI licensing regime risks repeating past mismanagement on a massive scale.
Second, CMOs represent only a fraction of India's creators. Emerging artists, independent authors, and practitioners of traditional arts often lack representation. The committee suggests a welfare fund for them, but such funds historically suffer from poor execution. A recent report noted a Labour Welfare Fund in Delhi held ₹5,200 crore in unspent money. Furthermore, royalty distribution prioritizing popular works raises doubts about benefits for lesser-known creators.
Third, the plan ignores that creators are also AI users. A Classplus report found 92% of Indian creators surveyed use generative AI. A blanket license would raise AI access costs as fees are passed to users, negatively impacting the very creators it aims to help, along with all other users.
Fourth, the global revenue focus punishes scale. It unfairly targets Indian AI startups with multinational ambitions, potentially charging them a tariff-like fee for exporting services. This creates an unclear and disadvantageous position for startups aiming to serve global markets from India.
A Crossroads for India's AI Ambitions
This proposal casts a shadow over India's upcoming AI summit, where the country seeks to position itself as an innovation leader. Conversations with entrepreneurs reveal they cannot afford the proposed licensing costs or the legal overhead for compliance. The likely result would be stifled domestic innovation or an exodus of talent to jurisdictions with more balanced copyright laws.
India now stands at a critical juncture. It can choose to adopt a forward-looking framework that encourages AI innovation and secures its place as a future powerhouse. Alternatively, it can cling to outdated regulatory models and watch this transformative opportunity migrate to other shores. The path chosen will define India's role in the global AI revolution.
