EU AI Board meets and adopts zero new binding rules

The European AI Board, which brings together member-state representatives, met on September 17, 2026 under the Irish Presidency of the EU Council — with Moldova attending as an observer for the first time. The agenda covered Commission enforcement priorities, transparency measures, market-surveillance cooperation, and frontier-AI cyber testing.

The outcome was plain: no adopted rule, no new company obligation, no revised compliance date. The AI Act timeline therefore stands as before — transparency obligations became enforceable on August 2, 2026, marking and detection duties under Article 50(2) apply from December 2, 2026 to systems already on the market, Annex III high-risk rules take effect December 2, 2027, and high-risk systems embedded in regulated products face compliance from August 2, 2028.

Work in progress leans toward coordination rather than legislation: recommendations to support AI literacy, market-surveillance governance arrangements, a possible secondment programme between national authorities and the European Data Protection Supervisor, and cybersecurity/AI testing infrastructure built with the EU Agency for Cybersecurity and the Joint Research Centre.

For companies in regulated sectors — aerospace, energy, industry, luxury — this is not a neutral pause. It extends a period where the compliance calendar is fixed but the precise enforcement mechanics remain under discussion, which bears directly on how conservatively firms should size the compliance investment they commit to now.