RBI says it is fully prepared for Anthropic's frontier AI model Mythos

The Reserve Bank of India (RBI) has said it is fully prepared to deal with Anthropic's frontier AI model Mythos even as it awaits final details on the project. The regulator has also issued advisories to banks and other regulated entities on dealing with cyber security threats that could arise from Mythos.

"To the best of my knowledge, access to Mythos has not yet been given. There are discussions on Claude and hopefully we will get it, we are fully prepared to deal with Mythos," Governor Sanjay Malhotra said at the post-monetary policy press conference.

Deputy Governor J Swaminathan said that the matter drew the attention of both the government and through the financial sector inter-regulatory forum.


"India has been included as one of the countries that will be part of this project, with select corporates and financial entities having access, but the details are fully awaited," Swaminathan said. "Once the opportunity opens on how to make use of it, in consultation with the government and other regulators, we will take further steps."

In the interim, the RBI has issued advisories to banks and other regulated entities on preparedness for AI-related developments, while also reinforcing cybersecurity readiness.

"We are mindfully prepared for dealing with cybersecurity threats of this nature as well as conventional ones. We will keep the market informed when we have full details," Swaminathan said.

Mythos is Anthropic's most advanced frontier AI model, currently being made available to a select group of trusted organisations and governments globally as part of the company's Project Glasswing.

The project provides restricted access to the highly controversial and powerful Claude Mythos AI model, which excels at finding zero-day software vulnerabilities and autonomously chaining them into working exploits.