The Algorithmic Alibi

APX Framework and the Architecture of Human Accountability in Autonomous Artificial Intelligence Systems This article presents two central contributions. The first is diagnostic: the formulation of the Algorithmic Alibi, the structural condition in which automated systems produce legally and socially relevant effects at the same time that the chain of human accountability dissolves within the technical architectures that sustain them. Drawing on Science, Technology and Society scholarship (Crawford, 2021; Gillespie, 2014), the article demonstrates that this invisibility is not incidental; it is structural. The second contribution is institutional: the APX Framework, a Decisional Infrastructure architecture grounded in Miguel Reale's tridimensional theory of law and structured around three pillars — Algorithmic Awareness, Strategic Intentionality, and Structured Accountability — which formalizes the Autonomy Manager and the Principle of Human Precedence as mechanisms for reconstituting accountability in automated environments. The paper further develops a stratified theory of civil liability and proposes the Algorithmic Governance Aptitude Index (IAGA) as a metric of organizational maturity. The APX Framework does not replace the EU AI Act, ISO/IEC 42001, or the NIST AI RMF. It grounds them, providing the accountability substrate these instruments presuppose but do not institutionalize.