The European Union's AI Act entered into force on Aug. 2, 2026 [1], granting Brussels the power to evaluate and sanction large AI models [1].
This regulatory shift marks the first comprehensive attempt by a major global power to impose legal accountability on the developers of powerful AI systems. By establishing a framework for investigation and penalties, the EU aims to ensure transparency and safety in an era of rapid technological expansion.
The legislation allows the EU to control and sanction large-scale models to address societal concerns regarding uncontrolled AI [1]. These measures are designed to ensure that the most influential systems remain under oversight to prevent systemic risks [1].
However, the feasibility of such regulation remains a point of contention among experts. In a discussion aired Aug. 3, 2026, Amélie Cordier said that the focus of regulation should be on the users rather than the software. "On n’encadre pas un marteau. C’est aux humains qu’il faut s’intéresser," Cordier said, suggesting that a tool like AI is not the primary object that requires framing.
Other critics suggest that the technology has already evolved beyond the reach of traditional legislation. Ethan Mollick said that the rapid proliferation of the technology makes total control impossible. "Le génie est sorti de la bouteille et on ne peut pas l’arrêter," Mollick said [2].
While the EU moves forward with its enforcement powers, some observers argue that the perceived risks of AI have not yet materialized in a way that justifies heavy regulation [2]. Despite these doubts, the provisions of the AI Act are now active, creating a legal mandate for AI developers operating within the EU to comply with new transparency, and safety standards [2].
“The EU can now control and sanction large AI models.”
The implementation of the AI Act establishes a global precedent for 'Brussels Effect' regulation, where EU standards often become the default for international companies. By shifting AI from a self-regulated frontier to a legally sanctioned industry, the EU is testing whether government oversight can keep pace with exponential technological growth or if the nature of AI renders traditional borders and laws obsolete.
