Note N02. Published on 24 September 2026. Theme: AI and digital. Lens: Is Europe losing, or choosing?.
The AI Act and business: legal uncertainty is now the second barrier to AI
Europe chose to be first to regulate artificial intelligence. Among companies that consider using it but do not, 54% now say the legal consequences are unclear, up from 46% in 2023.
The AI Act is the world's first general law on artificial intelligence. It entered into force on 1 August 2024 and applies in stages: prohibited practices since February 2025, and obligations for large general-purpose models since August 2025. The obligations for high-risk systems have been postponed: a regulation in force since 27 July 2026 moves them from August 2026 to December 2027 and, for AI built into regulated products, from August 2027 to August 2028, because technical standards and national authorities were not ready.
Is regulation holding companies back? Eurostat asks those that considered using AI but did not why. Lack of expertise is still the main reason (70%). But unclear legal consequences have risen from 46% to 54% in two years and are now the second reason, almost level with data protection (53%).
Lack of expertise is still the main barrier to AI; legal uncertainty is now the second
Reasons for not using AI, as % of enterprises that considered using it but did not, 2025
Loading the chart
The rise is widespread: legal uncertainty weighs more than in 2023 in 17 of the 18 countries with data for both years. In Spain it reaches 51%, and in Germany, which has no 2023 figure, 62%. The survey does not distinguish between laws, so not all of it can be attributed to the AI Act: data protection rules and intellectual property also count. But the 2026 postponement acknowledges the problem: complying with the law without finished technical standards was hard.
Legal uncertainty holds back more companies than in 2023
Enterprises not using AI because of unclear legal consequences, as % of those that considered it, 2023 and 2025
Loading the chart