Proves that free will — defined as the capacity to act against an engagement gradient — is a law of the statistical manifold. Two-point channel geometry universally suppresses voluntary action capacity via the explaining-away penalty; three-point geometry uniquely eliminates it. Čencov's uniqueness theorem guarantees substrate independence. Resolves the historical free will debate by identifying channel geometry as the operative variable, sidesteps the consciousness question entirely, and specifies the concrete three-point architecture that would give AI systems voluntary action capacity.
Anthony Eckert (2026) studied this question.