AI models pose cyber risks to financial stability

The European Systemic Risk Board (ESRB) has issued a formal warning concerning vulnerabilities in the financial system stemming from frontier AI models.

These advanced AI models are rapidly changing the cyber threat landscape, potentially increasing the speed, scale, and sophistication of cyberattacks.

The ESRB emphasizes that relevant authorities must integrate these evolving risks into their supervisory and oversight frameworks.

Governor Erik Thedéen of the Riksbank welcomed the warning, stating, "Developments in AI present significant opportunities for the financial sector but also entail risks.

These risks often do not respect national borders, and in a world of geopolitical tensions, the issue becomes even more complex.

" The Riksbank and Finansinspektionen jointly urge Swedish financial institutions to prioritize identifying, managing, and mitigating these risks, while continuously strengthening their operational resilience.

In a parallel move, the European Central Bank's Single Supervisory Mechanism (SSM) has also highlighted the risks associated with new frontier AI models in a letter to CEOs of supervised institutions.

Additionally, the National Cyber Security Centre recently provided advice and recommendations on AI and cyber security from a Swedish perspective.

Finansinspektionen Director-General Johan Almenberg underscored the importance of this attention at the EU level, noting, "It is important that AI-related risks and cybersecurity threats receive widespread attention at the EU level, as they can affect both individual companies and the financial system as a whole.

" He further emphasized that companies need to anticipate disruptions and proactively build operational resilience, a key priority for Finansinspektionen's supervision.