Reuters said major UK lenders still lack a timeline for access to Anthropic’s cybersecurity-focused model, while Bloomberg reported Bank of America is using Mythos internally for system testing amid operational and cyber risk questions.
Limited access to Anthropic’s Mythos AI model has intensified debate over Britain’s reliance on U.S. technology providers and highlighted how large banks are assessing advanced AI tools in controlled settings. Harriet Rees, Starling Bank’s chief information officer and the UK finance ministry’s appointed “AI champion,” said Britain needs domestic AI infrastructure, models and skills as major UK lenders still lack access to Mythos, which banks view as especially strong at identifying cybersecurity vulnerabilities. Separately, Bloomberg reported Bank of America is among Wall Street firms allowed to use the unreleased model internally for system testing, and CEO Brian Moynihan flagged risks around workload shifts, faster-moving vulnerabilities and the speed of institutional response.