Central banks are grappling with a new form of systemic risk driven by artificial intelligence, marking a significant shift from traditional financial challenges such as energy shocks, interest-rate fluctuations, and conventional banking failures. Recent developments indicate that AI has evolved from a supportive tool into a force capable of altering market behavior and exposing vulnerabilities within the global financial system.

The Bank of England has taken a notably proactive approach by initiating scenario analyses and simulations aimed at understanding how AI could impact financial stability. Sarah Breeden, the deputy governor for financial stability, highlighted the bank’s collaborative efforts with international counterparts to study the influence of AI "agents" on market dynamics, particularly in situations where herd-like behaviors could lead to rapid and disorderly sell-offs.

This heightened focus on AI comes amid criticism directed at the UK Treasury Committee for delays in extending the "critical third-party" regulatory framework to encompass AI and cloud-computing firms. Critics argue that these delays leave the financial system exposed, especially as banks increasingly depend on AI-driven models for risk management and decision-making.

In the United States, worries escalated following the emergence of "Mythos," a model developed by Anthropic and revealed in early April. Mythos possesses the capability to identify and exploit security vulnerabilities in operating systems and web browsers, raising concerns among regulators and financial leaders. Treasury Secretary Scott Bessent and Federal Reserve Chair Jerome Powell responded by convening an emergency meeting with major bank executives to address potential risks.

Complicating the landscape are geopolitical tensions, particularly between the US and China. Nvidia CEO Jensen Huang noted that despite US export restrictions on advanced AI technologies, China continues to advance its computing and semiconductor industries independently. This divide threatens to create parallel technological ecosystems and could undermine international collaboration essential for developing safer AI models. Without such cooperation, AI risks becoming a component of escalating technological and geopolitical confrontation, with significant repercussions for global financial stability.

Simultaneously, competition among technology firms intensifies the risk environment. Shortly after Anthropic introduced Mythos, OpenAI launched "GPT-5.4-Cyber," a model designed to detect system vulnerabilities. This rapid advancement in both offensive and defensive AI capabilities accelerates systemic risk, raising concerns that future financial crises may arise not from traditional economic factors like bad loans or inflation, but from algorithm-driven disruptions capable of widespread and instantaneous impact.