The Advisory Scientific Committee (ASC) has published a report examining how artificial intelligence could create or amplify systemic risks in the financial system.
The analysis reviews 11 key features of AI and how they interact with core sources of systemic risk, such as liquidity mismatches, common exposures, interconnectedness, lack of substitutability and leverage.
According to the ASC, five AI-related features pose the greatest concern: market concentration and entry barriers, model uniformity, monitoring difficulties, overreliance on and excessive trust in AI systems, and the speed at which AI operates. These factors could significantly magnify vulnerabilities across the financial sector.
While AI offers major benefits – including processing large volumes of unstructured data and improving risk management – the report stresses that a careful and balanced policy response is needed to mitigate risks. The ASC proposes combining competition and consumer protection policies with targeted changes to prudential regulation. These include adjustments to capital and liquidity rules, requirements around skin-in-the-game and technical sophistication, and enhanced supervisory frameworks.
The committee concludes that if authorities can keep pace with the fast evolution of AI, episodes of financial instability could become less frequent. Because AI systems transcend national borders, international regulatory cooperation will be essential.
