AI Governance in India Focuses on Safety and Innovation
New Delhi – In a statement to NDTV Profit, India's Electronics and IT Secretary S. Krishnan outlined the country's approach to regulating artificial intelligence. The strategy seeks a middle ground that protects citizens and data while preserving the momentum of technological breakthroughs.
Globally, AI models are becoming more powerful, and governments are scrambling to find effective ways to oversee them without stifling growth. India’s AI Mission is simultaneously expanding public cloud resources, fostering indigenous AI development, and boosting the overall AI ecosystem.
Krishnan pointed out that the world is moving toward a shared belief that AI must be regulated. The United States, once a vocal opponent of AI regulation, now imposes controls on cutting‑edge models, signalling a shift in policy attitudes.
India is conducting an in‑depth review of AI governance frameworks from different regions to craft a set of rules that are both practical and future‑ready. The goal is to establish robust safety measures and clear guidelines for AI platforms, without making the regulatory environment overly restrictive for innovators.
He acknowledged the difficulty of writing AI legislation due to the rapid pace of change. The law must evolve in tandem with technology, a nuanced balance that will require careful deliberation and time.
The secretary also highlighted India’s active participation in international AI governance forums, including discussions with the European Union and the PAX SILICA initiative led by the United States. Topics under review include AI infrastructure, computing resource access, and responsible AI development.
India has already signaled that it will not simply replicate the EU’s comprehensive AI Act nor adopt the early, less stringent US model. Instead, policymakers are designing a tailored framework that addresses the risks of frontier AI while granting sufficient flexibility for startups, research institutions, and innovation‑driven enterprises.