India may begin work on a dedicated legal framework for artificial intelligence (AI) as the technology continues to evolve, Secretary of the Ministry of Electronics and Information Technology (MeitY) S Krishnan said on Friday. Speaking at a CII conference, Krishnan indicated that the government now believes the time may be right to consider separate legislation to regulate AI.
Discussions on AI Legislation Underway
Responding to a reporter's question on AI regulations, Krishnan said discussions on a dedicated legal framework had commenced. The government had so far relied on existing legal provisions to address emerging challenges such as deepfakes and content labelling.
"My minister and I, both of us, have been on record earlier that we will look at AI regulation when the time is right and it appears that the time is getting right and we will start looking at it," Krishnan said.
Existing Laws Used for AI Concerns
Krishnan noted that the government has used the IT Rules and other provisions of existing law to address various concerns raised by AI. However, he added that the time has come to consider a separate legislation for AI regulation.
"We have used the IT Rules, we have used other provisions of existing law to address various concerns that AI raises, but now probably the time has come to look at a separate legislation to see how this is done," he said.
Timeline for Draft AI Legislation
When asked about the expected timeline for draft AI legislation, Krishnan said officials could prepare the legislative proposal but could not indicate when such a law might be introduced. "You see, the ministry and, at the official level, what we can do is prepare draft legislation. When it finally comes out is not something which I can comment on, especially when it is legislation," he said.
Export Curbs on Anthropic's Mythos Model Eased
On access to advanced AI models, Krishnan said export restrictions on Anthropic's Mythos model had been eased. However, he clarified that restrictions continue on Anthropic's next-generation models, including Fable 5 and Mythos 5.
"What they have restricted is on Fable 5. Fable 5 and Mythos 5? Mythos 5 is still restricted to their GlassWing partners. And the GlassWing partners are the initial lot which are US companies. Anthropic wants to extend it to entities in 15 countries, which includes India," Krishnan said.
US Government Clearance Still Required
Krishnan added that extending access to entities in the 15 countries, including India, would require approval from the US government. "Anthropic wants to extend it to entities in 15 countries, which includes India. But that still requires US government clearance. That is in process currently," he said.
Global Context of AI Regulation
The remarks come as governments across the world are exploring regulatory frameworks to address the rapid advancement of artificial intelligence, amid growing concerns over misinformation, deepfakes, algorithmic transparency, privacy, copyright and accountability.
