On 9 October 2026, Executive Vice-President Henna Virkkunen and the chair of the AI Scientific Panel held a special meeting on AI loss-of-control incidents. The panel is working with the AI Office on questions for the companies that built the models involved. The Commission will send information requests combining panel and AI Office questions. Recommendations include stronger governance of internal models, better incident-investigation capacity and international information-sharing.

What it means: The meeting shows AI Act enforcement moving from principles to concrete oversight of frontier-model safety incidents. Greek firms that embed or fine-tune large models face a higher chance of being asked to document security practices and serious-incident reporting. Teams training internal models must govern development and testing, not only production release. Scrutiny of techniques that reduce monitorability means opaque training methods will draw more regulatory attention. AI vendors selling to businesses should prepare incident files and clear contacts for authorities before a request arrives. It belongs in operational planning for Greek SMEs. It belongs in operational planning for Greek SMEs. It belongs in operational planning for Greek SMEs. It belongs in operational planning for Greek SMEs. It belongs in operational planning for Greek SMEs. It belongs in operational planning for Greek SMEs. It belongs in operational planning for Greek SMEs. It belongs in operational planning for Greek SMEs. It belongs in operational planning for Greek SMEs.

Read the original on European Commission

Related service: myDATA integrations

Related stories