The EU AI Act Enters Its Enforcement Phase
The European Union’s Artificial Intelligence Act, the world’s first comprehensive AI regulatory framework, has officially entered its enforcement phase in 2026. After years of legislative debate and a phased rollout that began in 2024, the Act’s most significant provisions are now binding on companies operating within the EU market.
The enforcement phase brings with it stringent requirements for high-risk AI systems, including mandatory conformity assessments, transparency obligations, and human oversight mechanisms. Companies deploying AI in critical sectors such as healthcare, transportation, and law enforcement must now demonstrate compliance or face penalties of up to 7% of global annual turnover.
What the Enforcement Phase Means for Businesses
For businesses operating in or selling into the European market, the enforcement deadline represents a watershed moment. The Act classifies AI systems into four risk categories — unacceptable, high, limited, and minimal — with corresponding regulatory burdens. Systems deemed to pose unacceptable risks, such as social scoring by governments and real-time biometric surveillance in public spaces, are now outright banned.
High-risk systems, which include AI used in recruitment, credit scoring, and medical devices, must undergo third-party auditing before deployment. Companies are scrambling to establish internal AI governance boards, conduct risk assessments, and document their training data sources — a process that many have described as more demanding than GDPR compliance was in 2018.
Global Ripple Effects
The Brussels effect is already visible. Several jurisdictions, including Canada, Brazil, and South Korea, have modeled their own AI legislation on the EU framework. Even in the United States, where federal AI regulation remains fragmented, individual states are introducing bills that mirror key provisions of the EU AI Act.
Major technology companies including Google, Microsoft, and OpenAI have publicly committed to aligning their AI development practices with the EU’s requirements globally, rather than maintaining separate compliance regimes for different markets. This de facto harmonization underscores the EU’s position as the world’s most influential technology regulator.
Challenges Ahead
Despite the Act’s ambitious scope, significant challenges remain. The definition of “high-risk” AI continues to be debated, particularly as foundation models and general-purpose AI systems blur traditional risk categories. National supervisory authorities across the 27 member states vary in their technical expertise and enforcement capacity, raising concerns about inconsistent application.
Small and medium-sized enterprises have voiced particular concern about the compliance burden, with industry groups warning that the cost of conformity assessments could stifle European AI innovation just as global competition intensifies. The European Commission has responded with an SME support package, including regulatory sandboxes and reduced fees for startups.
As the enforcement phase unfolds throughout 2026, the EU AI Act will serve as a real-world laboratory for AI governance — one that the rest of the world will be watching closely.
The Netherlands, for internationals
One email a week: the week's news for internationals in the Netherlands, new practical guides and what changed in the cities. No spam, unsubscribe any time.
