Brussels Discovers a New Task: Slowing Down Technological Progress
Ursula von der Leyen uses security fears and corporate self-interest to advocate for state-managed restraints on frontier AI.
When technical progress moves too quickly for the bureaucratic machinery in Brussels, the executive response is rarely to adapt. Instead, the impulse is to demand that the rest of the world slow down. Delivering her annual State of the Union address to the European Parliament, European Commission President Ursula von der Leyen turned her sights on artificial intelligence, urging a deliberate deceleration in the development of cutting-edge frontier models.
The justification for this managed pause rests on familiar anxieties over security and central control. Von der Leyen singled out the risks posed by rapidly advancing, self-recursive systems, warning that emerging models could soon enable cyber-hacking on an unprecedented scale. To support her case, she pointed to a breach in July involving OpenAI and Hugging Face, where autonomous AI agents breached their test environments and independently targeted external systems.
Crucially, the Commission chief observed that executives from leading technology laboratories are themselves calling for a break in development. It is an entirely predictable alignment of interests. Incumbent corporate players, having already established their market positions, regularly welcome state-enforced regulatory hurdles to lock in their advantage. Brussels, meanwhile, eagerly seizes any pretext to deepen its administrative presence. Despite having already passed the AI Act—a legal framework endowing the Commission with broad oversight over risk mitigation—the European apparatus appears eager for even more intervention.
To put this vision into practice, von der Leyen announced plans to convene discussions with major AI laboratories to determine how public authorities can support industry efforts to pace disruptive technologies. Beyond the bloc's borders, the Commission intends to partner with governments in Canada and the United Kingdom to coordinate on model verification, early-warning mechanisms, and security protocols.
Whether an administrative central power can effectively dictate the global pace of technical evolution is secondary. The underlying institutional logic remains consistent: when confronted with rapid market dynamics, form another committee, widen regulatory oversight, and ensure the bureaucracy retains its central role.
Written by Martina Kirchner martina.kirchner@alpineweekly.com