India's approach to artificial intelligence is characteristically bold and characteristically complex. The IndiaAI Mission, backed by a ₹10,372 crore outlay over five years, aims to build sovereign compute infrastructure, create indigenous large language models trained on Indian languages, and deploy AI tools across agriculture, health, and education at a scale no other country has attempted.

The compute pillar alone — 10,000 GPUs to be made available to startups and researchers at subsidised rates — has generated genuine excitement in the technology community. "Access to compute was the bottleneck. If the government fixes that, the next GPT-scale model might well come from Bengaluru or Hyderabad," said one AI researcher at IIT Delhi.

But critics point to a governance gap. India has no dedicated AI law, no independent AI regulator, and an evolving data protection framework that is still being operationalised. The tension between speed of deployment and adequacy of oversight is not unique to India — but the scale of ambition here makes the stakes unusually high.