ECB's Lagarde urges European AI autonomy over financial stability risks

Technological dependency on foreign providers

Speaking at the tenth annual conference of the European Systemic Risk Board (ESRB) in Frankfurt, European Central Bank President Christine Lagarde argued that Europe must build its own artificial intelligence infrastructure to safeguard financial stability. The development of frontier artificial intelligence tools remains concentrated in the United States and China, including models such as OpenAI's ChatGPT, Google's Gemini, Anthropic's Claude, xAI's Grok, and DeepSeek systems. Lagarde stated that relying entirely on foreign vendors leaves European financial institutions exposed if access is curtailed by overseas authorities.

It is vital that financial resilience does not depend on a switch controlled elsewhere.

Export restrictions test European access

Geopolitical dependency materialized as an active risk in June 2026, when an export control measure implemented by the United States abruptly halted European access to select advanced artificial intelligence systems. While access to public-tier versions resumed several weeks later, the most sensitive model tiers remained restricted to entities authorized by Washington. Lagarde noted that access restrictions were previously treated as hypothetical scenarios, but the summer restrictions demonstrated that European entities face concrete supply disruptions under changing foreign trade policies.

Autonomous algorithms and market vulnerabilities

Financial institutions across the euro area are rapidly deploying artificial intelligence into daily operations. Nearly 90% of significant banks in the euro area currently use artificial intelligence tools, while 70% of respondents in a survey of European Union securities market firms report plans to increase their spending on these technologies. Investment firms have deployed algorithmic models for years to parse data and evaluate portfolio risks, but the transition toward autonomous agents introduces new vulnerabilities. Lagarde noted that one of the largest hedge funds in the world has launched an aggressive strategy managed by an autonomous system. Unsupervised models could execute trades in ways their human operators never planned and cannot detect, amplifying cyber threats, stock market volatility, and price manipulation risks.

Taken individually, each risk is serious. But together they can interact and worsen across the entire financial system.

Systemic oversight and global cooperation

Lagarde recalled that the ESRB was created 15 years ago in the wake of the global financial crisis and the European sovereign debt crisis, bringing central bankers and supervisors together to monitor risk across the entire financial architecture. Citing physicist Stephen Hawking's warning that the future is a race between growing technological power and the wisdom with which it is used, Lagarde stated that artificial intelligence is now testing that supervisory model. To limit systemic vulnerabilities, Lagarde urged rival geopolitical powers to establish shared oversight standards, preventing high-capability models from falling into malicious hands while ensuring that authorities defending critical infrastructure retain full access.

I truly hope that such international cooperation will emerge.