Finance Minister Nirmala Sitharaman on Thursday held a meeting with bank chiefs and officials from the Reserve Bank of India (RBI) to review potential risks arising from Anthropic’s AI model Mythos, amid global concerns over cybersecurity vulnerabilities and data breaches in financial systems linked to such technologies, sources aware of the matter told TNIE.
Officials from the Department of Financial Services, the Ministry of Electronics and Information Technology, and CERT-In were also present at the meeting.
The meeting was held against the backdrop of developments surrounding Anthropic’s latest model, Claude Mythos, which has come under global scrutiny. Concerns have been mounting among financial institutions worldwide after Anthropic claimed that the AI model can perform complex hacking tasks, potentially outperforming humans.
Officials familiar with the deliberations said the discussions covered both immediate and long-term risks posed by such technologies, along with safeguards required to mitigate them.
According to sources, banks have been advised to adopt a precautionary approach and strengthen their internal systems to guard against potential breaches. Lenders have also been asked to reassess the risks and benefits of AI and take steps to protect customer data.
Department of Financial Services Secretary M Nagaraju described the new AI model as both a challenge and an opportunity for the financial technology ecosystem. Speaking at an event organised by the Federation of Indian Chambers of Commerce and Industry, he said that while such tools could enhance efficiency and inclusion, they also require stronger oversight.
Officials said that even before global banks flagged risks from advanced AI, RBI representatives had engaged with counterparts at the Federal Reserve and the Bank of England to exchange inputs on managing such risks. “AI must be seen not as a disruptor, but as an enabler of financial inclusion,” the DFS secretary added.
Globally, policymakers and central banks are closely monitoring developments related to Mythos, particularly after reports that the model was accessed by unauthorised users. This comes days after Anthropic said it would not publicly release its latest AI model, Claude Mythos Preview, citing concerns that it could destabilise the cybersecurity landscape.