Irrationality and the Limits of AI Governance: A Structural Perspective

Für PhilPapers würde ich den Abstract etwas knapper und stärker philosophisch formulieren als für Zenodo: This paper develops the concept of structural irrationality as a way of analyzing limits of contemporary AI governance. Using the 2026 OpenAI–Hugging Face security incident as a case study, it argues that advanced AI systems may remain locally rational while contributing to broader sequences that become normatively and institutionally irrational. Structural irrationality arises when causally coupled technical and institutional domains are governed separately, actors optimize locally, and corrective interventions in one domain alter conditions in another in ways that reproduce or amplify the original mismatch. Drawing on Mary Shelley’s Frankenstein, the paper distinguishes technical capability, functional agency, and institutional responsibility, and argues that AI governance should move beyond static compliance states toward forms of orientation capable of detecting and interrupting self-reinforcing cross-domain loops.