Trusted Execution is not a claim about a model. It is an evidenced property of an outcome produced through a governed effect path.
A system reaches Trusted Execution when the proposed consequence was evaluated against current authority and evidence, the decision was enforced before effect, and a receipt preserves what was evaluated, decided and executed or refused.
Trusted Execution = current authorization + enforced effect path + replayable evidence.
It does not mean the model is universally trustworthy, that prior approval remains valid forever, or that observability alone is sufficient. It is bounded to the evidenced execution instance.
Governed Effect Path · Execution Authorization Infrastructure