The signal this week is a regulatory deferral that boards are at risk of misreading. The European Union’s Omnibus amendment has postponed the AI Act’s most demanding high-risk obligations: stand-alone Annex III systems now face a December 2027 compliance date, and AI embedded in regulated products moves to August 2028. The headline reads as relief. A board looking for a reason to defer its own AI oversight architecture now has a date to point to.
The implication for the board is the opposite of the headline. The deferral moved a compliance deadline. It did not move the fiduciary obligation, because the obligation was never created by the AI Act. It was created by Delaware’s duty of oversight, which is indifferent to Brussels and unaffected by any European timetable. The general applicability and transparency provisions of the AI Act still take effect on 2 August 2026, and the duty to know what AI is doing inside the enterprise predates every statute and survives every postponement.
The convergent signal runs against the deferral. The SEC Investor Advisory Committee recommended in December 2025 that companies disclose how their boards oversee AI. APRA, in April 2026, told regulated boards that acknowledging the framework is not the same as operationalizing it. Three jurisdictions, one demand: show the system, not the intention. No Delaware court has yet adjudicated an AI-specific Caremark claim, which is exactly why documented oversight built now is the cheapest insurance a board will ever buy.
The executive takeaway is to treat the eighteen-month European reprieve as time to build the oversight architecture, not permission to defer it. The board that reads a scheduling change as a substantive one will be the board explaining, later, why it waited. The full board-level analysis sits in Touch Stone Publishers’ Scaling the AI Studio research hub.