A new mapping exercise will assist Finansinspektionen (FI) in better understanding AI application in the financial sector, including its use cases, risks, opportunities, challenges, and future avenues.
The initiative will also examine firms’ views on the EU AI Act and the impact that the new rules have on their businesses.
“AI is now everyday life in the financial sector. Through this exercise, we will obtain a better picture of how companies are approaching AI. It will also help us to better direct our resources to where they are most useful,” said Gustav Hagberg, AI specialist and coordinator of FI’s AI supervision.
Methodology and earlier findings
FI, including its Innovation Centre, will engage in an in-depth analysis that will largely be based on a survey sent to a broad range of entities under the agency’s supervision, including banks, insurance companies, and payment institutions.
FI examined the use of AI in the Swedish financial sector back in 2024, following a similar method. The findings, released at the end of 2024, revealed that from the 234 firms that responded to the survey, 84% of employees use AI tools such as Chat GPT, Copilot, or Gemini.
At the time, AI integration in firms’ own IT systems was lower given the higher thresholds needed. This could explain the large usage of public models. From the surveyed firms, 22% stated that they have AI systems in development, while 46% were engaged in pilot projects.
Most of these firms did not have a formally approved policy for the use or development of AI. In parallel, 91% of the firms had begun preparing for AI Act implementation, with FI stating that the financial sector will increase its AI use afterward.
The new mapping initiative, complemented by EU AI Act questions, comes after a sharp rise in AI use, including in the financial sector, which has in turn opened space for discussions around the opportunities that the technology brings, but also the risks it entails. One of the most prominent dangers lies in the area of cybersecurity.
AI concerns
Recently, the Swedish National Cybersecurity Centre (SNCC) published recommendations for companies to implement to ensure better protection against cybersecurity risks that may stem from advanced AI models. At the same time, the Central Bank’s latest Financial Stability Report revealed an increased risk of AI-enabled cyberattacks, pointing out the AI models’ unparalleled capabilities to detect vulnerabilities.
Internationally, a recent warning by the European Systemic Risk Board (ESRB) about vulnerabilities in the financial system driven by AI models has also raised the alarm.
“It is important that AI-related risks and cyber threats are widely recognized at the EU level, as they can affect both individual companies and the financial system as a whole. Companies need to assume that disruptions will occur and therefore build operational resilience. This is also a priority area in FI’s supervision,” said Finansinspektionen’s director General Johan Almenberg at the time.