European Commission convenes AI safety panel after loss-of-control investigations

The European Commission is holding a special meeting of its Scientific Panel on AI on October 9, 2026, to receive recommendations on frontier AI safety and security risks. Henna Virkkunen, the Commission’s executive vice-president for tech sovereignty, security and democracy, is scheduled to attend the meeting.

Incident investigations lead to developer questions

The panel has been investigating recent loss-of-control incidents involving AI models. Together with the Commission’s AI Office, it has worked on a set of questions for the companies that developed those models. The questions are directed at the developers behind the models involved, rather than AI companies generally.

The Commission did not identify the companies, models or incidents in its announcement. It used the term “loss-of-control” without describing what happened in the cases under investigation.

Virkkunen tied the need for scientific advice to the EU’s legal framework. She said the EU has the world’s first law addressing systemic risk from AI and needs state-of-the-art scientific input to support that work.

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, European Commission executive vice-president for tech sovereignty, security and democracy

A scientific remit across borders

The Scientific Panel brings together 60 independent experts. Its advisory role serves both the EU AI Office and national authorities, with responsibilities covering risks, the classification and assessment of models, and oversight of markets across national borders. The Commission lists four areas in its remit:

• Systemic risks: the panel provides scientific advice on this category of AI risk to the authorities it supports.
• Model classification: its work includes advice on how AI models are classified, or placed into categories.
• Evaluation methodologies: it advises on the methods used to assess AI models, not just the risks themselves.
• Cross-border market surveillance: its remit also includes advice on monitoring markets across borders, alongside its model-focused responsibilities.