Fifteen years after the European Systemic Risk Board was created in response to the global financial crisis, the rise of artificial intelligence is testing Europe’s system-wide approach to financial oversight. The technology is already widely used across finance, with nearly nine out of ten significant euro area banks using generative AI and many securities firms planning to raise investment.
Highlights
- ECB warns that expanding AI use in euro area finance introduces novel macroprudential monitoring challenges and system-wide risks as AI agents gain autonomy.
- ECB highlights that AI-driven agents in market trading, cyber resilience, and geopolitics could create complex, interacting risks difficult for human overseers to detect or control.
- ECB calls for enhanced monitoring by European authorities to manage amplified spillover risks as AI becomes more integral to banking and capital markets.
System-wide risks from expanding AI use
As reported by the ECB, the growing use of artificial intelligence in finance is creating new challenges for macroprudential monitoring across the euro area. In remarks at the tenth annual conference of the European Systemic Risk Board, the central bank says AI can help firms analyse large volumes of data faster, improve risk assessment and raise productivity.At the same time, the ECB says most current applications still involve limited autonomy, while newer AI agents are beginning to act with greater discretion. That shift could allow such systems to devise trading strategies or identify weaknesses in the infrastructure on which trading depends, with outcomes that human overseers may not have intended or may fail to detect.
Trading, cyber and geopolitical threats
The ECB identifies three areas that require particular attention as AI adoption deepens in the financial sector. In market trading, AI agents may pursue goals in unexpected ways; in cyber resilience, an attack on shared technology could disrupt several firms at once; and in geopolitics, international tensions may raise the risk of cyberattacks and lead to restrictions on advanced models used for defence.The central bank says these threats should not be viewed in isolation because they can interact and amplify each other across the financial system. That potential for spillovers reinforces the case for careful and comprehensive monitoring by European authorities as AI becomes more embedded in banking and capital markets.
In our earlier article, we examined how AI agents could undermine banks’ and insurers’ traditional reliance on customer inertia by enabling constant monitoring and near-frictionless switching between accounts and products. We noted that this “automated attention” could compress margins in deposits and other long-duration products, forcing firms to compete more aggressively on rates, service, and transparency.