Government Warns Banks to Brace for AI Cyber Risks

Canadian authorities, through the Office of the Superintendent of Financial Institutions (OSFI), have issued a specific warning to the local banking sector regarding the cyber risks associated with the advanced AI model Claude Mythos, developed by Anthropic. The move follows similar actions taken by other countries, including Japan and the United Kingdom.

“Advanced AI models, such as Anthropic Claude Mythos, shorten the time available for effective risk mitigation,” OSFI stated in an email sent to the financial sector, including banks, as reported by Reuters on Tuesday.

OSFI added that the bulletin provides guidance and outlines practices that institutions can adopt to improve the speed and effectiveness of risk identification, mitigation, and response. The email was also distributed to chief technology officers, chief information security officers, and chief risk officers across the financial industry, as well as to insurance companies.

Speaking to Reuters, OSFI explained that it maintains a technology-neutral approach while focusing on emerging technology risks, including those posed by advanced AI models like Mythos. “Our focus is not the technology itself, but how federally regulated financial institutions govern and manage the risks of its use,” the regulator said.

Discussions regarding the risks of Mythos were held in early April, attended by bank executives in Canada and regulators. The talks occurred shortly after US Treasury Secretary Scott Bessent and Federal Reserve Chair Jerome Powell issued stern warnings to bank CEOs about the risks associated with the AI model.

Meanwhile, several Canadian banks have been ramping up investments in artificial intelligence as they transition from experimental AI projects to practical applications in chatbots, building internal tools, and reducing reliance on third-party solutions.