Artificial Intelligence Governance of Clinical Algorithm Post-Market Surveillance Under Unmeasured National Regulator Enforcement Dormancy: A Healthcare Political Hazard Estimation

National regulators are formally responsible for post-market surveillance of clinical artificial intelligence (AI) algorithms, yet in many low- and middle-income countries enforcement activity is unmeasured and may be dormant. The framework combines a principal-agent model of regulator-industry interaction with a hazard function in which the probability of detecting algorithm failure depends on unobserved enforcement intensity. Using a panel of 24 African countries from 2000 to 2023 (N = 560 country-years), the analysis estimates the relationship between national income, electricity access, urbanisation and regulatory capacity, drawing on World Bank World Development Indicators. The ordinary least squares estimates show that access to electricity and urban population share are positively associated with GDP per capita, while the coefficient on the time trend is negative, with an R-squared of 0.591. Enforcement dormancy is not a simple absence of activity but a structured condition that redistributes risk toward patients in low-capacity settings. The contribution is a formalisation of post-market surveillance as a political hazard problem and a set of propositions linking unmeasured enforcement to algorithm harm. The conclusion draws out implications for transparency, regional regulatory pooling and the design of surveillance obligations that remain robust when national enforcement cannot be verified.