Existing governance frameworks produce evidence that controls operated. They do not establish whether those controls remain correct under adversarial conditions or conflicting institutional incentives. This paper identifies that limitation as a structural feature of current assurance regimes and introduces adversarial sufficiency as a new class of governance evidence standard. The standard requires demonstrable correctness under defined adversarial conditions, external to the governed entity, and validated through falsifiable evidence artefacts. A four-domain validation methodology and a cross-layer incentive inversion test are presented to operationalise the standard. A V1.0 implementation demonstrates that adversarial sufficiency can be established through manual execution with documented results, prior to automation. The result is not a refinement of existing frameworks, but a definition of the evidentiary boundary they do not address.
Barry et al. (Tue,) studied this question.