Capacity enhancement for more efficient AI regulation

Important results from the project

The project partially achieved our original goals. The most important goals that were met were promoting general understanding of how market surveillance authorities view AI issues, where they stand, and how they can use AI and handle AI oversight. We conducted a GAP analysis that focused particularly on testing. We built networks and knowledge that will support Sweden´s AI regulation work going forward.

Expected long term effects

The project is expected to have several long-term effects for Sweden´s AI regulation work. The networks and collaborations we built with the government investigation and various authorities will continue after the project ends. The knowledge and understanding developed about AI oversight and testing will support Sweden´s implementation of the AI Act. Overall, the project has created a foundation for more informed and effective AI regulation work in Sweden.

Approach and implementation

The project didn´t develop quite as intended, but we redesigned it while working to be as helpful as possible for future processes. We could have focused more on understanding how AI products work. The early workshops on market surveillance and digital AI products needed to be much longer. This was too small a project to handle that well. We underestimated how difficult it would be for researchers and authorities to understand each other´s areas.