If your highest-risk AI model degraded this morning, who would own it, and by when?
Governance Decision Latency is the time from an AI system's automatic suspension to the next accountable human decision. It is the one governance number that cannot be produced by writing documents. Answer three questions for a transparent estimate of yours, and where it sits against the closed-loop threshold.
The estimate: your committee interval's average wait, adjusted for how drift is routed, and gated by whether anything acts automatically. The formula is on this page; nothing is hidden.
A forwardable summary for your board packet, plus the free templates to close the loop. No sales sequence.
This audit is an educational planning aid based on the answers you provide. It is not legal, clinical, or regulatory advice, and it is not a determination about any specific institution's liability or negligence. Governance Decision Latency is defined in full at /standards/decision-latency/, as a stage of Closed-Loop Defensibility.