The head of the Bank of England’s Prudential Regulation Authority (PRA) has warned of “quite significant disruption” to financial services from new AI tools, including Anthropic’s Mythos and ChatGPT 5.5 Instant, reported Reuters.
Sam Woods, who leads the PRA, said these systems are becoming better at finding weak points, increasing pressure on banks to fix them quickly.
He described this patching work as “the main driver of outages” across the financial system.
Speaking at UK Finance’s Growth Delivery Summit, Woods said firms would have to strengthen routine cyber hygiene and react more quickly, with AI-based defensive tools set to play a larger role.
Anthropic introduced Mythos to a small group of businesses in April. Cybersecurity specialists see the model as a challenge for banks and older technology infrastructure, though a cyber group co-led by the Bank of England concluded last month that the sector was ready for those issues.
Previously, Reuters has reported that Anthropic was preparing to offer Mythos to banks in Europe, as lenders around the world examine the technology after major US banks were given early access.
One person said Anthropic plans to extend access to Mythos to banks in Europe and the UK, along with other organisations.
Last month, regulators in both the US and the UK held discussions with large banks about cyber security concerns linked to Anthropic PBC’s newest AI model.
In Britain, the Bank of England, the Financial Conduct Authority and HM Treasury were in urgent contact with the National Cyber Security Centre and were also in discussions with the country’s largest banks, the Financial Times said.
Last month, the Financial Times also reported that the UK government was considering a shared testing framework for general-purpose AI systems used by lenders, following concerns raised by the Bank of England last year over how those models are being evaluated.
