Just days after regulators scrambled to warn global banks about the dangers of Anthropic’s new AI model (Mythos), the technology is already crossing the Atlantic.
On April 16, Anthropic confirmed it will grant UK financial institutions controlled access to Mythos within the next week. The move marks the latest expansion of Project Glasswing, the company’s tightly gated program designed to arm defenders before the offensive capabilities of advanced AI spill into the wild.
Mythos, internally tested by Anthropic, has stunned experts with its ability to spot and chain zero-day vulnerabilities across every major web browser and operating system. As one Polymarket commentator put it, the model appears “intelligent enough to spot weaknesses in almost every computer on earth.”
Our analysis published on April 12 captured the mounting anxiety. Banks, with their complex patchwork of legacy systems and modern platforms, sit at the heart of the risk. Regulators on both sides of the Atlantic have been in crisis mode: US Treasury Secretary Scott Bessent and Fed Chair Jerome Powell summoned Wall Street CEOs for urgent briefings, while Germany’s BSI and the UK’s Bank of England, Treasury, FCA, and National Cyber Security Centre have all convened high-level discussions.
Anthropic insists the transitional period is “fraught” but ultimately positive. CEO Dario Amodei has praised the early Glasswing partners, which already include JPMorgan Chase, AWS, Google, Microsoft, and others, for confronting the threat “head-on.” The company’s own statement was blunt: “Advanced language models are here. The threat is not hypothetical.”
For UK banks, the arrival of Mythos next week brings a double-edged sword. In the right hands, it could become the ultimate red-team tool, uncovering flaws before malicious actors do. In the wrong ones, it could accelerate the very cyberattacks that keep chief information security officers awake at night.
Pip White, Anthropic’s head for the UK, Ireland and northern Europe, noted significant engagement from British bank CEOs in recent days. The message from the City is clear: the age of AI-native cybersecurity has arrived, ready or not.
White confirmed the timeline bluntly: “That is in the very near term, in the next week.”
For British banks, Mythos is arriving whether they feel ready or not. It is a powerful new tool that could dramatically strengthen defences, or accelerate the very threats they fear most.
The age of AI-native cybersecurity has officially begun.
See Also:
Is Anthropic’s Mythos Able to Hack the Global Banking System? | Disruption Banking
