India's government is exploring a significant new step in tech regulation: requiring pre-approval for advanced AI models before their public release. The move signals a global push for AI safety, but what could it mean for India's tech future?
What Is Being Proposed?
The Government
of India is considering a new legal framework specifically for artificial intelligence that could require developers of the most powerful AI systems to undergo safety checks before releasing their products to the public. This represents a potential shift from India's previous 'light-touch' stance, which relied on existing laws and voluntary guidelines. Officials are examining measures like pre-release testing, evaluations in controlled 'sandbox' environments, and clear assignment of responsibility for any harm caused by these advanced models. The focus is on what are being called 'frontier models'—cutting-edge AI whose capabilities could pose significant risks to society, including enabling sophisticated cyberattacks or generating highly convincing deepfakes. A formal consultation paper is expected to be released in the coming weeks to gather feedback from the industry and public.
Why the Change in Approach?
This move toward stricter regulation comes as policymakers acknowledge that the risks associated with AI have 'crossed a threshold'. The government's concerns are driven by the rapid advancement of AI capabilities and their potential for misuse. Issues like the spread of misinformation through deepfakes, financial fraud, impersonation, and the increasing difficulty in distinguishing real content from AI-generated material are key drivers behind the proposal. There is a growing sense that existing laws, like the Information Technology Act of 2000, are no longer sufficient to address the unique challenges posed by modern AI. Recent amendments to the IT Rules in early 2026, which mandate the labelling of 'synthetically generated information', were a first step, but the new proposal suggests a more comprehensive approach is now seen as necessary.
The Core Debate: Safety vs. Innovation
The central challenge for the government is to strike the right balance between ensuring public safety and fostering innovation in a sector critical to India's economic ambitions. A key part of the debate is how to define a 'high-risk' or 'frontier' model, a distinction that would allow regulators to apply strict scrutiny only where it's most needed, without burdening the entire AI ecosystem. Tech industry experts and startups have voiced concerns that an overly restrictive pre-approval regime could stifle research, slow down development, and put Indian companies at a competitive disadvantage globally. An earlier government advisory in March that suggested 'explicit permission' was needed for even unreliable AI models faced significant pushback from startups, forcing a clarification that it was aimed at large platforms. The upcoming consultation will be crucial for navigating this tension, ensuring that regulation doesn't become a bottleneck for legitimate and beneficial AI development.
How This Fits into the Global Context
India's move is part of a worldwide trend as governments grapple with how to govern powerful AI technologies. The European Union has its landmark AI Act, which uses a risk-based approach, and the United States has pursued a combination of executive orders and voluntary commitments from major tech firms. India's proposed framework appears to be drawing lessons from these international approaches, particularly the US focus on safeguards for high_risk systems. However, experts in India have urged the government to develop a framework suited to India’s unique scale, languages, and application_led AI market rather than simply copying models from elsewhere. The goal is to align with global principles of responsible AI, such as those from the OECD, which India has endorsed, while creating rules that work for the domestic ecosystem.
