EU AI Act Enforcement Begins: New Compliance Requirements Take Effect

European Union Implements Landmark AI Regulation Framework

The European Union's Artificial Intelligence Act, one of the world's most comprehensive AI regulatory frameworks, has officially entered its enforcement phase as of September 2025. The legislation, which came into force in August 2024, establishes a risk-based approach to artificial intelligence governance across all 27 member states.

Risk-Based Classification System

The AI Act categorizes AI systems into four distinct risk levels: unacceptable risk, high risk, limited risk, and minimal risk. This represents a paradigm shift in how we approach AI governance, stated European Commission spokesperson Maria Schmidt. We're moving from reactive regulation to proactive risk management.

Applications deemed to pose unacceptable risks are completely banned, including AI systems that manipulate human behavior or use real-time remote biometric identification in public spaces. High-risk applications, particularly those used in healthcare, education, and critical infrastructure, must undergo rigorous conformity assessments and maintain ongoing compliance monitoring.

General-Purpose AI Requirements

The legislation specifically addresses foundation models like ChatGPT, requiring transparency obligations for most systems and additional evaluations for high-capability models. Open-source models receive reduced requirements, explained AI policy expert Dr. Thomas Weber, but all providers must ensure their systems meet fundamental rights protections.

The Act creates a European Artificial Intelligence Board to promote cooperation between national authorities and ensure consistent enforcement across the EU. Like the GDPR, the AI Act has extraterritorial reach, applying to providers outside the EU if they serve users within the bloc.

Implementation Timeline

Provisions will phase in over the next 6 to 36 months, giving organizations time to adapt their AI systems to the new requirements. Companies developing or deploying AI within the EU must now conduct thorough risk assessments and implement appropriate safeguards.

For more information on the AI Act's provisions, visit the European Commission's official page.