The Overton Framework: Protective Computing in Conditions of Human Vulnerability

Contemporary software architecture is implicitly governed by the Stability Assumption: the belief that users operate under continuous connectivity, cognitive surplus, environmental safety, and institutional trust. This assumption is increasingly incompatible with real-world conditions in which human–computer interaction occurs during crisis, displacement, illness, coercion, or systemic instability. This work introduces the Overton Framework, establishing Protective Computing as a systems-engineering discipline oriented toward safety, containment, reversibility, and essential utility rather than engagement or growth. The framework formalizes Stability Bias as a function of dependency, vulnerability, and irreversibility; defines a Vulnerability State Machine for crisis-aware interaction; and proposes normative protective design principles governing local authority, exposure minimization, reversibility, and coercion resistance. An accompanying reference implementation demonstrates technical feasibility within a privacy-first, offline-capable architecture, supporting empirical evaluation through a Protective Legitimacy Score derived from reversibility, exposure, local authority, cognitive load, and sovereignty metrics. Together, the Canon and Implementation Companion articulate both the theoretical foundation and operational evidence model required to evaluate software designed for conditions of human vulnerability. Protective Computing reframes digital system design as a safety-critical ethical obligation, proposing a new baseline for trustworthy software in the twenty-first century.