Notes and Briefs.

Architectural commitment under the framing of pilot, experiment, or deployment-grade trial. AI capabilities are evolving faster than the contemporaneous record of why specific architectural choices were made. By the time the next administration arrives and asks for the structural reasoning, the operative governing logic has either been preserved or has not.

Most institutional records identify the binding moment with the formal approval. The structural binding usually occurred earlier — sometimes weeks earlier — in a conversation that was not minuted. By the time the formal decision arrives, the structural binding has often already occurred.

The handover packet records what was decided. It rarely records why each material assumption was held, what trade-offs were accepted as the binding condition, who held authority to do the trading, and how the irreversibility was structurally classified at the moment of binding.

An executive transitioning out of a role is rarely binding commitments that operate on their own time horizon. The structural condition that determines whether the transition holds is not the quality of the handover packet — it is the gap between the outgoing actor's operational horizon and the successor's inheritance horizon. Successor intelligibility (Doctrine § IX) applies most directly here.

Architectural commitments to specific AI vendors are being made at signing under language that describes them as pilots or experiments. The structural condition reads differently from the contractual framing: the binding moment is at signing, and the gap between the signing room's operational horizon (12–18 months) and the architectural horizon (4–7 years) produces a recurring class of unstructured commitment.