AI and systemic risk are presented as a balanced assessment of how rapidly adopted, large-scale AI models can both strengthen and destabilize the financial system. The piece argues that while AI promises productivity gains and better decision-making, its interaction with core sources of systemic risk demands targeted competition, consumer protection, and prudential policy responses.

Key takeaways:

1. AI’s current productivity contribution appears modest at the aggregate level, but its impact is highly uneven across occupations, with some clerical and routine roles particularly exposed to automation while many professional roles are augmented rather than replaced.

2. AI can amplify existing systemic risks in finance—such as liquidity mismatches, common exposures, interconnectedness, lack of substitutability, and leverage—through features like model uniformity, opacity, speed, concentration of providers, overreliance, and susceptibility to hallucinations and misinformation.

3. The authors call for a mix of strengthened competition and consumer protection policies, along with adjustments to macroprudential regulation and supervision (including capital and liquidity requirements, transparency and labelling of AI use, “skin in the game” standards, and enhanced supervisory capacity), as well as international coordination to prevent cross-border fragilities.

“AI’s ability to process immense quantities of unstructured data and interact naturally with users allows it to both complement and substitute for human tasks. However, using these tools comes with risks,” says Lumsdaine.

Read the article.