EX ANTE Ω Authorization Before Action On the Validity and Procedural Admissibility of Intelligent Systems.

This work introduces a doctrine of ex ante procedence validity for intelligent systems, establishing authorization as a condition of admissibility prior to execution. It formalizes a strict separation between technical capability and legitimacy, asserting that no action may be considered valid solely by virtue of performance, optimization, or outcomes. The doctrine defines procedural states of authorization, revocation, and nullity, and articulates an external, model-agnostic precondition layer that determines whether execution may proceed at all. Rather than proposing governance, alignment, or safety mechanisms, this framework operates as a prior validity filter, preserving interruptibility, responsibility, and the possibility of refusal under conditions of increasing autonomy and scale.