Finance leaders warn over Mythos as UK banks prepare to use powerful Anthropic AI tool

UK banks are set to gain access to Anthropic’s latest AI model, Claude Mythos, within the next week, despite warnings from finance leaders about its potential risks. The model, currently limited to select US companies such as Amazon, Apple, and Microsoft, has raised concerns due to its advanced capability to identify and exploit vulnerabilities in IT systems. Anthropic has described Mythos as posing an unprecedented threat, with implications for economic stability, public safety, and national security. Anthropic’s head of UK, Ireland, and northern Europe operations, Pip White, highlighted significant interest from UK financial sector CEOs ahead of the rollout. The company has emphasized that AI models like Mythos now surpass most human experts in coding and cybersecurity skills, increasing the risk of cyberattacks and system breaches. This has prompted discussions among finance ministers, regulators, and executives at the IMF and World Bank spring meetings in Washington, where the broader geopolitical context, including tensions from the US-Israeli conflict with Iran, also featured prominently. Canadian Finance Minister François-Philippe Champagne underscored the challenge of managing an “unknown unknown” threat, stressing the need for robust safeguards to protect financial system resilience. Meanwhile, Andrew Bailey, governor of the Bank of England and chair of the Financial Stability Board, acknowledged the rapid pace of AI development and the difficulty regulators face in determining the appropriate timing and extent of regulatory intervention. He noted the tension between fostering innovation and ensuring security, questioning when to establish “rules of the road” for AI technologies. The imminent introduction of Claude Mythos to UK banks marks a critical moment in balancing the benefits of advanced AI with the risks it presents to financial infrastructure. As institutions prepare to deploy the tool, ongoing dialogue among global regulators and policymakers will be essential to mitigate potential fallout while harnessing AI’s economic potential.