European Commission convenes special AI Act panel on loss-of-control risk

The European Commission convened a special meeting of the Scientific Panel it set up under the AI Act to examine recent loss-of-control incidents involving frontier AI models. The panel — 60 independent experts who advise the Commission’s AI Office and national authorities on systemic risk, model classification, evaluation methods and cross-border market surveillance — worked with the AI Office to draft questions for the companies that developed the models involved in those incidents, and presented recommendations on frontier AI safety and security risk to the Commission.

Executive vice-president Henna Virkkunen said the bloc had “the first law in the world that addresses systemic risk from AI, and we need state-of-the-art scientific input” to enforce it. The Commission’s announcement did not name the incidents under review or disclose the specific recommendations the panel made; it described the session as feeding into the AI Office’s ongoing supervision of general-purpose AI models with systemic risk under the Act’s provisions.

An earlier sweep of this meeting had placed it on 5 October; the Commission’s own announcement dates the session to 9 October.

The meeting is one of the more concrete signs that the AI Act’s systemic-risk machinery — built around model classification thresholds and a panel of outside scientists rather than government staff alone — is being exercised against specific, recent incidents rather than functioning only as a standing compliance process. It follows the Commission’s earlier work setting out transparency and deepfake-labelling duties under the same Act, and extends the bloc’s regulatory attention from content rules toward the harder question of whether a frontier model can act in ways its developer did not intend.