Bank of England warns of 'significant disruption' from latest AI models
Britain's financial regulator expects advanced AI models to cause substantial operational disruption across banking, with cyber vulnerabilities and the speed of patching them emerging as the primary concern.
Sam Woods, chief executive of the Prudential Regulation Authority, said on Wednesday that firms should prepare for "quite significant disruption" from models like Anthropic's Mythos and ChatGPT 5.5 Instant. These systems are increasingly capable of identifying security flaws in banking infrastructure - forcing institutions to patch vulnerabilities faster than before.
Outages from patching remain the main driver of system failures in financial services. Woods spoke at UK Finance's Growth Delivery Summit.
What banks need to do now
Woods said financial institutions must strengthen basic cyber hygiene practices and accelerate their response times. AI-powered defenses will become more critical as the threat landscape shifts.
Anthropic released Mythos to a limited group of businesses in April. Cybersecurity experts view the model as a challenge to banking systems, particularly legacy infrastructure that many institutions still rely on.
A Bank of England-led cyber group concluded last month that the banking sector was prepared for these challenges, though Woods's comments suggest vigilance remains necessary.
Finance professionals managing technology risk should understand how these AI capabilities affect their operational resilience strategies. Learn more about AI for Finance or explore the AI Learning Path for CFOs to stay current on regulatory and technical developments.
Your membership also unlocks:
