Anthropic's latest frontier AI model, known as Claude Mythos, is being feared by the global banking sector. Financial regulators in the U.K. are moving to evaluate the potential threats posed by the powerful AI tool, with officials from the Bank of England, Financial Conduct Authority, and HM Treasury reportedly having critical discussions with the National Cyber Security Centre. At the same time, major British banks financial bodies are expected to receive details within the next two weeks on the risks associated with Claude Mythos.
Their concern mirrors actions taken in the U.S., where Treasury Secretary Scott Bessent recently convened leaders from prominent Wall Street banks for discussions around Claude Mythos. Similarly, executives and regulators from Canadian banks held a meeting last Friday to examine the dangers linked to the AI model.
Why Banks Are Running Scared Of Anthropic's Claude Mythos
Anthropic last week announced Project Glasswing, a cybersecurity initiative aimed at strengthening defences across essential digital infrastructure, including banking. Anthropic has given select partners — including Amazon Web Services, Google, Apple, Microsoft, Broadcom, Nvidia, CrowdStrike, Cisco, JPMorgan Chase, Palo Alto Networks, and the Linux Foundation — along with around 40 additional organisations controlled access to test and apply Claude Mythos Preview.
But why so?
These partners are using Claude Mythos Preview to scan for and patch vulnerabilities in their own systems and in widely used open-source projects before any harmful exploitation can occur.
Across the world, banks and financial institutions rely on the infrastructure provided by these tech majors for their operations. With vast amounts of financial assets and critical customer data in their domain, they are especially worried, because they could be prime targets for cyberattacks if Claude Mythos falls into the wrong hands.
Already, Anthropic has admitted that during internal evaluations, the model identified thousands of high-severity vulnerabilities, including zero-day exploits in major operating systems and web browsers — some of which had remained undetected for decades. Safe to assume global banking operations are being run on such platforms as well, and they face potential risk.
The AI company has further cited the model's immense power and risks of misuse by malicious actors as reasons it hasn't been made public yet. Due to its ability to reveal hidden cybersecurity weaknesses in critical software systems, experts fear bad actors could use Claude Mythos to carry out large-scale attacks — potentially outpacing the response time of cybersecurity teams.
With Project Glasswing, tech majors are hoping they might one-up cyber attackers. However, the underlying fear among the banking sector is clear: if misused by malicious actors, Claude Mythos could trigger a new era of cyber threats that are faster, bigger, and more complex — aptly termed “AI doomsday” by many.
Also read: Apple's AI Chief John Giannandrea To Exit Next Week
Essential Business Intelligence, Continuous LIVE TV, Sharp Market Insights, Practical Personal Finance Advice and Latest Stories — On NDTV Profit.