US Regulators Shift Stance on Anthropic's Mythos AI, Urge Banks to Adopt for Cybersecurity
What began as significant alarm over Anthropic's advanced new artificial intelligence model, Mythos, is now evolving into a posture of cautious acceptance and strategic deployment. According to a detailed report from Bloomberg, senior United States government officials are actively instructing major Wall Street financial institutions to implement the Mythos system within their internal operations. The primary objective is to substantially enhance their defensive capabilities against increasingly sophisticated cyber threats.
Major Banks Begin Testing Following High-Level Directive
The initiative follows a high-level meeting convened last week by Treasury Secretary Scott Bessent and Federal Reserve Chair Jerome Powell. Bank executives were summoned to Washington D.C., where officials underscored the critical importance of utilizing the Mythos AI to proactively identify and address system vulnerabilities. In response, JPMorgan Chase has already commenced testing the model. Other leading lenders, including Goldman Sachs, Citigroup, Bank of America, and Morgan Stanley, are expected to follow suit imminently.
This represents a notable strategic pivot. Previously, regulators had issued warnings about the potential risks associated with AI-enabled cyberattacks, even describing the Mythos model as potentially 'exceptionally dangerous'. The current directive, however, encourages banks to run the AI against their own systems not in reaction to an immediate crisis, but as a preemptive measure to build long-term resilience.
Controlled Access Through Project Glasswing
For the present, Anthropic is strictly limiting access to Mythos under a program named Project Glasswing. This initiative grants controlled access to a select group of partners, including Google, Microsoft, JPMorgan Chase, and cybersecurity firm CrowdStrike. The goal is to harness the model's formidable capabilities exclusively for defensive purposes within a secure and monitored environment.
Anthropic has repeatedly emphasized that an uncontrolled, widespread release of such powerful AI could have severe consequences for global economies, public safety, and national security. Cybersecurity analysts note that the company's measured approach reflects both genuine caution and its established reputation as a 'safety-first' organization within the AI industry.
Broader Push for Financial Sector Preparedness
National Economic Council Director Kevin Hassett articulated the urgency behind this move, framing it as part of a broader governmental push to ensure financial institutions are thoroughly prepared. "We're taking every step we can to make sure everybody is safe from these potential risks," Hassett stated, highlighting the proactive nature of the strategy.
International Ripples: Canada Holds Awareness Session
The concerns and discussions surrounding Mythos are not confined to the United States. Recently, executives from Canada's largest banks and top financial regulators gathered for a meeting organized by the Canadian Financial Sector Resiliency Group (CFRG). Chaired by Bank of Canada Chief Operating Officer Alexis Corbett, the session included representatives from the Department of Finance, the Office of the Superintendent of Financial Institutions (OSFI), and executives from Canada's six biggest banks plus Desjardins Group.
The meeting focused on the cybersecurity risks posed by Anthropic's new Claude Mythos AI model, amid growing international apprehension that such technology could be weaponized to exploit software weaknesses. A Bank of Canada spokesperson, Paul Badertscher, clarified that the gathering was not an emergency response but a 'situational awareness' session. "It can still hold meetings at the request of its members. 'Hey guys, we need to pay attention, there is something going on. Let's get together and talk about this.' That's what this was," Badertscher explained.
This coordinated response from North American financial regulators and institutions signals a critical transition in the narrative around powerful AI—from one dominated by fear to a more nuanced approach focused on controlled, defensive application to safeguard vital economic infrastructure.
