Ninth meeting of the AI Board: incidents involving the most advanced models, AI literacy recommendations and market surveillance

The European Artificial Intelligence Board (AI Board), composed of representatives of the Member States, met for the ninth time on 17 September 2026 – the first time since most of the AI Act became applicable on 2 August 2026. The meeting was chaired by the Irish Presidency of the Council of the EU and was attended for the first time by Moldova as an observer.

The Commission informed the Member States of its priorities and the steps taken so far in enforcing the AI Act and presented new capabilities of the most advanced AI models (frontier AI), including recent incidents.

On AI Act implementation, the Commission reported on the preparation of recommendations to support AI literacy – which under Article 4(3) are adopted by the AI Board itself – and on measures supporting the transparency rules of Article 50, namely the Code of Practice on Transparency of AI-Generated Content and the transparency guidelines.

The agenda also covered cooperation on market surveillance and the governance structures for pre-market conformity assessment.

The Commission also updated the Member States on the Action Plan on Cybersecurity and AI, in particular on building infrastructure for evaluating and testing the cybersecurity capabilities of the most advanced models.

Discussions further covered, among other things, strengthening AI capabilities, AI safety and AI uptake in key sectors.

The tenth meeting is scheduled for 18 November 2026 in Brussels, the day after the Apply AI Summit.