There have been mounting concerns about how robustly European Union member states may police artificial intelligence use since the AI Act took effect on Aug. 1. Opinions are divided on whether authorities will take a “light touch” approach to regulation–or hit companies with a sledgehammer for noncompliance. One thing’s for sure, though: The pace of AI innovation will make enforcement very difficult, experts said. 

Europe may be at the forefront of putting rules in place to prevent AI from causing harm to consumers, but it does not necessarily follow that regulators will have much appetite or capacity to enforce these rules or that companies know how to comply with them.

Part of the problem is that the act leaves it up to individual countries to decide whether they want to set up their own dedicated AI regulator to oversee use, or whether they want an existing authority–namely, data protection authorities (DPAs)–to add AI monitoring to their remit. Critics have pointed out that many DPAs are already bowing under the pressure of dealing with a growing number of complaints relating to the General Data Protection Regulation (GDPR).