'Strategic autonomy' in AI crucial, avoid overdependence on single source: IT Secretary, S Krishnan

New Delhi, India should pursue "strategic autonomy" by building enough domestic capability in key parts of the AI ecosystem and avoid overdependence on any single technology source or location, while remaining firmly integrated into trusted and verified global supply chains, IT Secretary S Krishnan said on Tuesday.

Krishnan said India has an opportunity to emerge as a global leader in building AI-powered applications and solutions for the world.

He said India should leverage inputs from multiple sources while avoiding overdependence on any single country or source, and build sufficient capabilities to become an indispensable part of the global AI value chain.


"Sovereign has a particular context. I think in the Indian context, even in a global context today, the appropriate term would be strategic autonomy. So, we need to have autonomy in the way in which we build these structures," Krishnan said.

He was speaking at the launch of an independent multimodal AI evaluation platform by AI4Bharat (research initiative at IIT Madras) and Josh Talks AI.

Krishnan further cautioned against overdependence on one source, or any one particular location.

"And equally, we need to have some capacity domestically in order to make sure, especially in a country the size of India...in some part of this value chain, you need to have a significant presence so that you can't be taken for granted, or people can't say we are going to bypass you completely," Krishnan said.

The IT secretary spoke in favour of resilient and diversified AI supply chains, acknowledging global interdependence, while avoiding excessive reliance on any single source or place.

"That is how you build interdependence and a structure where you can rely on trusted and verified sources for certain kinds of things, and you are also part of that trusted and verified supply chain with enough diversification, enough resilience so you can go through a number of situations," Krishnan said.

Speaking to reporters on the sidelines, Krishnan said the time has come to start thinking about an artificial intelligence law and the broader regulatory framework that should govern the technology, and added that the government is currently in the process of gathering feedback.

"I think AI law is what we mentioned...that some form of regulation of AI will have to come in at some point, and probably the time to start thinking about it is now. So we are, at this point, primarily looking at feedback," he said.

Asked about the status of draft IT rules tweaks, he said, "We'll have to see which elements we look at, in case we are looking at a new statute...then whether we should take it into an Act or whether we should do it only by rules... Those are the aspects."