• Government reviews cyber risks from advanced AI models like Claude Mythos.
• Banks asked to boost cyber defences, monitoring and threat sharing.
• Authorities say risks are unprecedented but systems remain secure.
India stepped up scrutiny of emerging artificial intelligence risks to the financial sector, with Finance Minister Nirmala Sitharaman chairing a high-level meeting with senior bankers and regulatory officials to assess potential cybersecurity threats posed by advanced AI models.
The discussions come amid growing global attention around Anthropic’s unreleased Claude Mythos model, which is reported to have advanced capabilities in identifying software vulnerabilities across operating systems and web browsers.
1 April 2026
Get the latest issue of Outlook Business
According to reports cited by The Hindu, Indian authorities are currently evaluating the extent of risks such technologies could pose to the domestic financial ecosystem.
The meeting included participation from senior officials of the Reserve Bank of India, NPCI, CERT-In, and top banking executives. The Ministry of Finance said in a post on X that banks have made progress in strengthening cybersecurity systems but warned that emerging AI-driven threats require “very high vigilance, preparedness and coordination.”
Unprecedented AI Cyber Threat
During the meeting, Sitharaman described the nature of emerging risks from advanced AI systems as “unprecedented,” stressing the need for stronger coordination between financial institutions and regulators. Banks were advised to enhance cyber monitoring frameworks and engage specialised cybersecurity experts to strengthen defensive systems.
Officials also discussed the importance of immediate reporting of suspicious cyber activity to CERT-In and other relevant authorities.
Push for Real-Time Threat Sharing
A key outcome of the discussions was the push for a real-time cyber threat intelligence sharing system between banks and security agencies to enable faster detection of risks and coordinated responses across the financial ecosystem. While officials noted that existing banking systems remain secure for now, the review reflects broader concerns around advanced AI-driven cyber risks.
The assessment draws on industry discussions around advanced AI systems, including Anthropic’s Claude Mythos model, which is said to outperform humans in certain cybersecurity tasks and detect large volumes of software vulnerabilities.
In an April 7 note titled “Assessing Claude Mythos Preview’s cybersecurity capabilities”, Anthropic said that “in the long run, we expect that defense capabilities will dominate, that the world will emerge more secure with software better hardened—in large part by code written by these models—but the transitional period will be fraught,” highlighting both long-term gains and near-term risks.
Authorities cautioned that such AI systems could also be misused in the near term to target digital infrastructure, including banking systems.
