India wants a sovereign AI model. It’s a big goal, and success means stringing together contributions from a range of entities, including Big Tech firms and domestic AI developers of all scales. The ideal version of this plan was meant to lead to plenty of competition—no preordained winners.

The latest development, however, is one that runs against that framework. Sarvam has been tapped to build India’s own large language model (LLM), while firms like Gan AI, Soket AI Labs, and Krutrim—all shortlisted to be part of the process—are still waiting for a call from the government.

Never one to let an opportunity slide, Sarvam is using its newfound status to raise up to $100 million in fresh venture funding, people in India’s AI ecosystem told The Ken.

As a nation, India is making a big bet to bring into existence a sovereign AI model. But it’s off to a rocky and familiar start: there’s one anointed winner while many others are left confused. 

Besides, there are questions about whether India’s flagship AI initiative is on the right course. Sarvam is a two-year-old company that’s funded with non-sovereign capital from the likes of Lightspeed, Peak XV, and Khosla Ventures. Does that warp the nature of India’s plan to create a sovereign AI model?

Abhirami explains in this week’s Make India Competitive Again, as read by Snigdha Sharma.

Read this as a newsletter: https://the-ken.com/newsletter/make-india-competitive-again/indias-ai-mission-needs-many-heroes-its-settled-for-one-sarvam/