Log In Get Started
Economy May 11, 2026 12:12 PM

Bank regulator warns latest AI models could trigger 'quite significant disruption' in financial services

Prudential Regulation Authority chief flags vulnerability discovery and fast patching needs as primary outage drivers; AI defences urged

By Leila Farooq

The head of the UK’s Prudential Regulation Authority warned that advanced generative AI models, including Anthropic's Mythos and ChatGPT 5.5 Instant, are likely to cause substantial disruption across financial services. Sam Woods said the models' enhanced capacity to detect system weaknesses, combined with the imperative for banks to patch swiftly - which he identified as a leading cause of outages - means firms must improve cyber hygiene and accelerate responses, with AI-based defences growing in importance. Woods delivered the remarks at UK Finance's Growth Delivery Summit. Anthropic introduced Mythos to a limited set of businesses in April; cybersecurity experts see the models as challenging for banks and legacy systems, though a Bank of England co-led cyber group concluded last month that the sector was prepared.

Bank regulator warns latest AI models could trigger 'quite significant disruption' in financial services

Key Points

  • The PRA head said it is "reasonable to expect quite significant disruption" to financial services from advanced AI models such as Anthropic's Mythos and ChatGPT 5.5 Instant.
  • Woods identified the growing ability of these models to find vulnerabilities, and banks' need to patch quickly - which he called "the main driver of outages" - as a risk for the financial system.
  • Firms are being told to improve basic cyber hygiene, speed up responses, and adopt AI-driven defences; Anthropic rolled out Mythos to a limited number of businesses in April, and a BoE co-led cyber group determined last month that the sector was prepared.

Sam Woods, chief executive of the Prudential Regulation Authority (PRA), said on Wednesday that it is "reasonable to expect quite significant disruption" to financial services from the latest generation of artificial intelligence models, naming Anthropic's Mythos and ChatGPT 5.5 Instant as examples.

Speaking at UK Finance's Growth Delivery Summit, Woods pointed to these models' improving ability to uncover vulnerabilities within systems. He said that the subsequent need for banks to fix those weaknesses rapidly - which he described as "the main driver of outages" in the financial system - raises the prospect of elevated operational stress.

Woods urged firms to strengthen basic cyber hygiene and to accelerate their incident response capabilities. He said that as threats evolve, "AI-driven defences" will become increasingly important in limiting disruption and protecting critical services.

Anthropic made its Mythos model available to a limited number of businesses in April. Cybersecurity experts have expressed concerns that models of this capability present novel challenges for the banking sector and for older, legacy technology stacks that underpin many institutions' operations.

At the same time, a Bank of England co-led cyber group assessed the sector's readiness and determined last month that financial firms were prepared for these challenges. Woods' comments underline a tension the regulator sees between the potential for disruptive acts enabled by advanced models and the industry's current resilience.

The PRA chief's remarks focused on concrete operational dynamics: faster detection of vulnerabilities by AI, the consequent pressure to deploy patches quickly, and the risks that rapid remediation itself can introduce outages. He called for reinforced hygiene basics as a first line of defence and for the adoption of AI-enabled protective measures to keep pace with developing threats.

Woods' address at the summit highlights the regulator's view that while some preparedness exists, the evolving capabilities of AI models mean that firms must continue to adapt processes, tools and response times to mitigate a rise in service disruptions.

Risks

  • Increased frequency or scale of operational outages due to rapid patching after vulnerabilities are discovered - impacts banking operations, payments, and market infrastructure.
  • Challenges posed by advanced AI models to legacy technology systems in banks, potentially stressing older infrastructure and elevating remediation complexity.
  • A gap between evolving AI-enabled threats and firms' response speeds or hygiene standards, which could expose financial services to higher disruption despite assessments of preparedness.

More from Economy

Trump Says He Will Discuss Iran with Xi in China Visit, But Says He Doesn’t Need Beijing’s Help May 12, 2026 U.S. Posts Reduced $215 Billion April Surplus as Refunds and Outlays Rise May 12, 2026 Bundesbank's Nagel Says Iran Conflict Could Force ECB to Raise Rates May 12, 2026 EIA Sees U.S. Electricity Use Climbing to New Highs Through 2027 May 12, 2026 Chicago Fed’s Goolsbee Flags Broadening Inflation, Cites Services as Main Concern May 12, 2026