The European Commission published on Wednesday, April 9, a roadmap aimed at making Europe a "leading AI continent." One of the five pillars of this strategy concerns "regulatory simplification," with measures presented as necessary to reduce the administrative burden on businesses. Although the Commission affirms its commitment to safe AI that respects fundamental rights, the Center for AI Safety (CeSIA) is concerned that the European institution is yielding to the narrative of unbridled competition and urges the Commission not to sacrifice the safety of European citizens in a global race for AI power.
By declaring that "the race for AI leadership is far from over" and making "regulatory simplification" a priority focus of its strategy, the Commission is embracing a narrative that undermines the essential regulatory and cooperative efforts needed to develop safe and reliable AI.
"Unregulated competition for the development of artificial intelligence constitutes one of the worst scenarios for protecting ourselves from the systemic risks of this technology. Questioning the regulatory framework that has barely been established undermines citizens' trust and sends a very bad signal about our continent's strategic independence by yielding to pressure from the United States. We urge the European Commission to maintain its initial course: prioritize robust and protective regulation rather than aligning with a narrative that sacrifices our safety on the altar of competitiveness. Europe has the opportunity to lead the way toward AI that truly benefits everyone, rather than engaging in a race whose consequences could be devastating for our societies," says Charbel Raphael Segerie, Executive Director of CeSIA.
Rhetoric Directly Borrowed from Technology Companies
CeSIA regrets that the European Commission has adopted a narrative promoted by technology companies, which claim to be more concerned about "regulatory uncertainty" than the systemic risks associated with insufficiently regulated AI. The publication of the action plan coincides with the launch of a public consultation aimed at "determining whether European digital legislation adequately addresses the needs and constraints of businesses" – a concerning inversion of priorities, where business requirements seem to take precedence over the primary objective of the law: ensuring that the common market serves the needs and interests of European citizens.
This framing is all the more problematic as it echoes the rhetoric recently adopted by the U.S. administration. During the AI Action Summit in Paris, U.S. Vice President J.D. Vance declared that he was "not here to talk about AI safety […] but about AI opportunity," asking that “deregulatory flavor making its way into a lot of the conversations”.
A Symbolic Retreat for AI Regulation
Caught in a storm of conflicting interests, the European Commission seems to be revising its AI priorities, stating through Henna Virkkunen, Commissioner Vice-President for Technological Sovereignty, that it would be "harmful to no one to reduce certain reporting obligations" for companies under the Artificial Intelligence Regulation (AI Act).
The Commission's official communication emphasizes the need to make the regulatory framework "streamlined and simple." By suggesting that certain requirements could be adjusted to meet competitiveness imperatives, it creates uncertainty about the real priority given to AI safety. While the AI Act Code of Practice has not yet been published, this approach blurs the political message carried by the European Union and gives the impression that rules can be renegotiated before they are even applied.
Europeans Demand Stronger Regulation, Not Deregulation
Seeing Europe adopt the logic of deregulation and competition sends a signal contrary to our continent's strategic autonomy and the clearly expressed desire of its population to firmly regulate this technology. Experts have been warning for years about the considerable risks of AI deployed without solid safety guarantees and the urgency of ending the unregulated race for AI power.
