The EU AI Act Enters a New Phase
The European Union’s Artificial Intelligence Act, the world’s first comprehensive AI regulatory framework, entered its next enforcement stage in July 2026. Businesses operating in or selling to the European market now face concrete compliance obligations that go far beyond the high-level principles discussed when the Act was first passed.
The AI Act classifies AI systems into four risk categories: unacceptable, high, limited, and minimal. As of this month, the provisions covering high-risk AI systems — including those used in critical infrastructure, education, employment, and law enforcement — are now fully enforceable across all 27 member states.
What Changed in July 2026
The most significant shift is the mandatory conformity assessment for high-risk AI systems. Companies must now document their training data sources, demonstrate bias testing results, and maintain human oversight protocols. Failure to comply carries fines of up to €35 million or 7% of global annual turnover — whichever is higher.
Transparency obligations also kicked in for limited-risk systems. Any AI that interacts directly with consumers — chatbots, recommendation engines, emotion-recognition systems — must clearly disclose its artificial nature. The days of users unknowingly speaking to an AI are over in Europe.
Impact on Dutch and European Businesses
For the Netherlands’ thriving tech sector, the impact is twofold. Larger firms with dedicated compliance teams have largely adapted, but small and medium enterprises face a steeper climb. The Dutch government has launched a subsidised AI compliance helpdesk through the Rijksinspectie Digitale Infrastructuur (RDI) to assist SMEs with the transition.
European startups are also watching closely. While the Act creates compliance costs, it also establishes a clear rulebook — something investors have long said was missing. Venture capital firms report increased confidence in European AI startups precisely because the regulatory uncertainty has been resolved.
What Experts Say
Privacy advocates have praised the enforcement timeline, noting that Europe is setting the global standard while other jurisdictions debate voluntary guidelines. Industry groups, however, warn that the compliance burden could slow innovation compared to less regulated markets.
For businesses still preparing, the message from Brussels is clear: the grace period is over. Companies that have not yet completed their AI system inventory and risk classification should treat July 2026 as the point of no return.
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.
