- Around half of the financial institutions surveyed in Switzerland use AI or are developing applications; a further 25% plan to use it within three years.
- Only half of the institutions have an explicit AI strategy; existing governance focuses on data protection, security, data quality and explainability.
- FINMA recommends early contact for AI use in critical processes and refers to Risk Monitor 2023 and Supervisory Communication 08/2024.
FINMA has published the results of a survey of around 400 financial institutions in Switzerland on the use of AI (Media release).
Around half of the institutions surveyed use AI or have applications in development, and a further 25% plan to use it in the next three years. On average, 5 applications are in use and 9 are under development. Larger companies use more applications than smaller ones, and smaller ones rely more heavily on externally developed applications.
Number of applications, cumulated across all respondents:
Only half of the institutions use AI as part of an explicit AI strategy. The Governance-frameworks – where they exist – primarily consider the topics of data protection, security, data management and enterprise risk management and – specifically for AI – the topics of data quality, data protection and explainability.
FINMA also draws attention to the Risk monitor 2023 and in particular the comments on outsourcing risks and their Regulatory Notice 08/2024 – Governance and risk management in the use of AI point out. It recommends contacting FINMA at an early stage if AI is to be used in critical processes or for the calculation of regulatory parameters.