Artificial intelligence (AI) models that allocate agricultural credit, inputs and extension services among smallholder cooperatives are increasingly audited by committees in which ministry observers sit alongside cooperative delegates. When those observers collude, the audit verdict can be corrupted even though every individual vote is honestly cast. the present analysis develops a governance integrity bound for Byzantine-fault-tolerant (BFT) consensus scoring applied to such audits, and evaluates the operating characteristics of the design through a replication-based power analysis of a before-after-control-impact (BACI) estimator. No observational data are used or claimed. The analysis characterises the design under a declared variance structure (between-cluster SD = 6.0, residual SD = 4.5, intraclass correlation = 0.64) across 2000 replications per design cell, with cluster counts of 8, 16, 24 and 32 and effect sizes of 3, 5 and 8 percentage points. The computed intraclass correlation of 0.64 implies that 52 clusters are required to detect a 5-percentage-point effect at 80% power. BFT consensus scoring converts observer collusion from a threat to verdict validity into a bounded, estimable design parameter, and it specifies the conditions under which the resulting integrity bound holds.
No takes yet. Share an insight, caveat, or question.
Fikru et al. (2026) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: