The Pulse
EU Panel Examines AI Loss-of-Control Incidents at Special Meeting
The European Commission says its Scientific Panel has been investigating recent AI loss-of-control incidents and prepared questions for the companies that developed the models involved. The panel is due to present recommendations on frontie
AI.info Team · Written with AI ·
Questions for companies, recommendations for the Commission
The European Commission’s Scientific Panel is examining recent AI “loss-of-control incidents” as the EU puts its AI Act to work on systemic risks. At a special meeting on October 9, 2026, the panel is due to present recommendations on frontier AI safety and security. The Commission says the panel has also worked with its AI Office on questions for the companies that developed the models involved.
The announcement does not name the models, describe the incidents, or publish the questions. It also does not say whether the companies have responded. That leaves the meeting’s recommendations as the clearest signal of what the Commission’s scientific advisers see as the risks requiring attention.
Virkkunen ties the panel to enforcement
Henna Virkkunen, the Commission’s executive vice-president for Tech Sovereignty, Security and Democracy, is attending. Her message links scientific advice to the EU’s legal authority and its ability to enforce rules on advanced AI.
“The EU has the first law in the world that addresses systemic risk from AI, and we need state-of-the-art scientific input. AI capabilities are advancing rapidly, but with a strong legal framework, decisive enforcement and the best scientific minds supporting this work, Europe can lead in ensuring safe and secure AI.”
Henna Virkkunen, Executive Vice-President for Tech Sovereignty, Security and Democracy, European Commission
The statement sets out the Commission’s position, not an independent assessment of how effective the law or its enforcement will be. The meeting’s practical significance will depend on what the panel recommends and how the Commission uses that advice.
A 60-member panel with a wide remit
The panel has 60 independent experts. It advises the EU AI Office and national authorities on systemic risks, classifying general-purpose AI models, evaluation methods and cross-border market surveillance.
The Commission’s panel page also lists expertise in capability evaluation, risk assessment, technical mitigation, misuse and deployment risks, cybersecurity and compute measurement. Members serve two-year terms in a personal capacity. Their remit makes the panel a technical adviser to regulators, rather than a separate enforcement authority.
The panel’s listed expertise also includes risks emerging at a systemic level and cyber-offence risks. Its members were selected from a range of countries, with the Commission’s panel page describing geographic and gender balance as considerations in the selection process. The page says most experts work in academia, while others come from public-sector-linked organisations, independent research or civil society. Their appointments are renewable, 24-month terms, and members serve in a personal capacity.
Incident details remain undisclosed
The Commission’s September account of an AI Board meeting said officials had discussed recent AI incidents and new frontier-model capabilities. The October announcement connects that attention to the Scientific Panel’s investigation of loss-of-control incidents, but offers no further incident details or findings.
For now, the Commission has announced a process: experts examined incidents, helped frame questions for model developers and are presenting recommendations. The published notice does not say what the recommendations are, or identify the systems they concern.