Autonomous and AI-driven systems are commonly evaluated through internal correctness, optimization quality, and component reliability. Yet empirical failures repeatedly emerge not from internal malfunction, but from breakdowns at system boundaries—where authority, coordination, and legitimacy intersect. This paper identifies boundary failure as the dominant risk in autonomous systems, arising when authority becomes ambiguous, coordination degrades, or silence is misinterpreted across operational domains. Building on prior work distinguishing resilience from uptime, treating silence as a valid operational state, and formalizing authority contraction and refusal as safety invariants, this paper names and characterizes the Boundary Authority Problem as a systemic failure pattern. It argues that without explicit boundary governance, autonomous systems tend to persist beyond legitimacy, optimize beyond authority, and act beyond coordination—producing failure modes that internal correctness alone cannot prevent. Autonomous system failure is reframed not as a technical deficiency, but as a governance failure at the boundaries of control. Whether autonomous systems can safely scale without explicit boundary authority remains an open question.
David Forbes (Thu,) studied this question.