Artificial Intelligence Governance of Snakebite Antivenom Stocking Models Under Unmeasured Regional Procurement Cartel Allocation: A Healthcare Political Robustness Evaluation

Antivenom stocking models increasingly inform subnational procurement decisions in snakebite-endemic health systems, yet their governance is complicated by a political condition that the models themselves cannot observe: regional procurement cartels that allocate antivenom volumes across districts through informal agreements among suppliers, brokers and political intermediaries. the present analysis evaluates the robustness of a before-after-control-impact (BACI) design for detecting the effect of AI-governed stocking interventions when cartel allocation operates as an unmeasured cluster-level confounder. No observational data are used or claimed; the analysis characterises the operating characteristics of the design under a declared variance structure with between-cluster standard deviation of 6.0, residual standard deviation of 4.5 and intraclass correlation of 0.64. Power was estimated across 2000 replications per design cell for cluster counts of 8, 16, 24 and 32 and effect sizes of 3, 5 and 8 percentage points. The design required 54 clusters to detect a 5 percentage point effect at 80 per cent power. The results indicate that the BACI estimator's power is highly sensitive to cluster count and that the cartel-allocation confounder inflates the minimum detectable effect beyond what conventional governance evaluations assume. AI governance of antivenom stocking must treat procurement cartel allocation as a structural feature of the evaluation problem rather than as a data limitation to be resolved by larger samples.