India's approach to artificial intelligence is maturing. The government is now preparing a dedicated legal framework for AI, signalling a significant shift from regulating online content to governing the core technology itself.
From Content Moderation to Core Technology
For years, India's digital
governance has been anchored by the Information Technology Act of 2000 and its subsequent rules, which primarily focus on the role of intermediaries—the platforms that host user content. This framework treats issues like misinformation and harmful content as a matter of platform responsibility. Recent amendments have targeted AI-generated 'deepfakes' by mandating quicker takedowns and content labelling. However, policymakers now believe these risks have crossed a threshold where simply managing content is no longer enough. The government's new thinking, reflected in recent statements by the Ministry of Electronics and Information Technology (MeitY), suggests a move to regulate the AI models themselves, not just the content they produce. This represents a fundamental change from a reactive posture to a proactive one.
What a Broader AI Law Could Involve
A dedicated AI law would look very different from existing IT rules. Instead of focusing on social media platforms, it would address the entire lifecycle of an AI system, from its development and training to its deployment in the real world. Officials have indicated that the planned legislation will address threats that conventional content moderation cannot, such as autonomous AI agents taking unauthorised actions, models identifying weaknesses in critical infrastructure, or generating sophisticated cyberattacks. The framework is also expected to prominently feature rules around deepfakes, financial fraud, and online impersonation. This new approach could establish a safety regime for the most powerful AI models before they are released to the public, a significant departure from India's largely hands-off strategy to date.
A 'Risk-Based' Regulatory Model
India appears to be drawing inspiration from global regulatory trends, particularly the European Union's AI Act. The EU's model uses a risk-based approach, categorising AI systems into tiers from minimal to unacceptable risk. 'High-risk' applications, such as those used in critical infrastructure, law enforcement, or employment, are subject to the strictest requirements, including mandatory conformity assessments. While India's final framework is still under discussion, the forthcoming consultation paper is expected to explore a similar risk-based calibration. This would mean that an AI used for a harmless creative task would face light regulation, while an AI used for credit scoring or medical diagnoses would face intense scrutiny. The proposed Digital India Act, which is expected to replace the old IT Act, has also contemplated regulating high-risk AI systems and defining 'no-go' areas for the technology.
The Human in the Loop
A central pillar of the proposed legislation is the principle of human oversight. As AI systems become more autonomous, regulators are focused on ensuring that a human remains accountable and in control. Potential requirements being examined include mandatory human approval for sensitive AI actions, restrictions on AI access to critical data, and the implementation of 'emergency stop' controls to halt unsafe AI behaviour. This aligns with guidelines recently proposed by India's Supreme Court for the use of AI in the judicial system, which also emphasize strict human oversight. The focus is on ensuring that technology remains a tool, with final decisions and accountability resting with people, not algorithms.
Balancing Innovation with Safety
The government's challenge is to draft a law that mitigates harm without stifling India's burgeoning AI industry. Over the past few years, the official stance has shifted several times, reflecting this difficult balance. After initially stating in 2023 that no AI-specific law was being considered, the rapid evolution of AI capabilities has forced a change in thinking. The IT Ministry has stressed that it wants to hear from all sectors—from banking and manufacturing to financial services—that are rapidly adopting AI. The upcoming consultation paper is intended to gather broad input from citizens, industry, and experts to build a consensus. The goal is to create a framework that fosters responsible innovation while protecting citizens from the novel risks that this powerful technology introduces.
