Governing frontier general-purpose AI in the public sector: adaptive risk management and policy capacity under uncertainty through 2030

This project is an academic or policy paper rather than a software tool. Its 'defensibility' is low because it is a public-facing intellectual contribution intended for open discourse, lacking a technical moat, proprietary data, or network effects. In the realm of AI policy, the 'competitors' are think tanks (e.g., Ada Lovelace Institute, Center for Security and Emerging Technology) and international bodies (OECD, UN). The low star count and lack of repository activity confirm this is a static document or a reference implementation of a framework. Frontier labs are unlikely to compete here as this addresses the *regulation* and *management* of their products by public entities, though they may lobby against the frameworks proposed. The displacement horizon is set to 1-2 years because the speed of AI development and the shifting regulatory landscape (e.g., EU AI Act implementation) quickly render specific governance prescriptions obsolete.