Bank of England Governor Andrew Bailey has called on global financial regulators to urgently assess the cybersecurity risks posed by advanced artificial intelligence systems, particularly Anthropic’s latest Mythos AI model. According to Bloomberg, Bailey warned that regulators need to rapidly evaluate how frontier AI models could expose vulnerabilities in banking infrastructure and critical financial systems. His comments reflect growing concern that AI is no longer just a productivity tool but also a potential systemic risk in financial stability.
A key issue is the model’s ability to identify hidden software vulnerabilities and weaknesses in digital systems, capabilities that could be used for both defense and malicious cyberattacks. Bailey noted that cyber threats have risen sharply in importance since the global financial crisis and described them as a risk that “never goes away.” For banks, this means AI could potentially accelerate sophisticated attacks on payment rails, trading systems, customer databases, and core banking infrastructure.
The article also points to a coordinated regulatory response in the UK. The Bank of England, Financial Conduct Authority (FCA), HM Treasury, and the National Cyber Security Centre (NCSC) are reportedly working together to assess the implications for the banking sector. Major banks, insurers, and financial exchanges are expected to be briefed on the risks, underscoring how seriously regulators are treating AI-enabled cyber threats.
Overall, the broader takeaway is that AI risk is moving into the realm of financial stability and systemic regulation. Bailey’s warning signals that regulators now see advanced AI models as a possible catalyst for next-generation cyberattacks against the banking system. This marks an important shift: AI oversight is becoming as much about resilience, operational security, and macro-financial risk as it is about innovation.
