European Union struggles to police fast-evolving AI
PUBLISHED May 21, 2026, 11:40 AM ET
Read, Watch or Listen
The European Commission this week released official guidelines defining high-risk artificial intelligence under the EU’s landmark AI Act, highlighting the difficulty of regulating rapidly evolving software with slow-moving legal tools. Enforcement of key provisions has already been delayed, with implementation now pushed to December 2027 and August 2028. Unlike physical products whose safety features are fixed at manufacture, advanced AI systems can change capabilities through updates and user interactions, complicating oversight. Regulators favor detailed, time-consuming verification, while technology companies seek rapid global deployment. Observers are watching whether authorities adopt automated testing tools to monitor AI systems in near real time.
By Daniel Hayes | JQJO News
Timeline of Events
- This week European Commission issues AI guidelines
- Recently lawmakers delay core AI Act enforcement
- December 2027 First major enforcement deadline set
- August 2028 Additional AI Act obligations take effect
- Ongoing AI systems evolve via continual software updates
- Ongoing Regulators rely on static legal definitions
- Future Regulators may adopt automated testing tools
- Future Real-time code checks could supplement oversight
News Intelligence
- AI is everywhere, from your phone to your car. As it evolves, so do its risks and benefits. The EU's struggle to regulate AI could impact your safety and privacy. Check your device settings regularly to manage AI features.
Comments