Why the EU rewrote its landmark AI law (www.theparliamentmagazine.eu)

In December 2023, European Commission President Ursula von der Leyen announced significant revisions to the EU Artificial Intelligence Act, transforming it from a values-driven framework to one more geared toward fostering innovation. The latest amendments, finalized on May 7, 2026, notably delay the implementation of high-risk AI system regulations to 2027-2028 and exempt many industrial applications, a move aimed at facilitating compliance for sectors critical to European economies. However, rules for general-purpose AI models developed by firms like OpenAI remain on track for implementation in August 2026, indicating that the regulatory urgency persists for certain AI applications. This shift in the EU's approach reflects a growing emphasis on competitiveness in the global AI landscape, prompted in part by lobbying from major U.S. tech companies. While supporters argue that the amendments were necessary to enhance Europe’s innovation landscape, critics contend that the changes primarily provide breathing room for companies without fundamentally altering the law’s ambitious risk-based framework. With the new regulations set to roll out soon, questions arise about the EU's capacity to adapt to rapid technological advancements while ensuring robust oversight of AI systems, marking a pivotal moment in the ongoing evolution of AI regulation in Europe.