India’s sovereign AI stack is already based on the American Cloud

In Part 1 of this series, I examined why India’s sovereign AI narrative risks conflating national origins with product-market fit. But the sovereignty argument has a problem that goes beyond market dynamics. There is an evidence problem.

The AI ​​Impact Summit in India last week produced a series of announcements about sovereign artificial intelligence. Sarvam AI has launched a 105 billion parameter model trained from scratch on Indian infrastructure. The Government of Odisha is committed to a sovereign AI data center. IIT Madras has laid the foundation for India’s first sovereign AI research park. The Union ministers spoke of technological independence. The word “sovereign” appeared in almost every keynote.

There is a structural contradiction in this formulation. India’s most sensitive citizen-facing digital platforms, the systems for processing biometric data, vaccination records, identity documents and government procurement, are already running in the American cloud…