Brokering permission under uncertainty: How regulatory sandboxes architect innovation governance

Financial technology innovations challenge regulatory governance by destabilizing the classification schemes, risk models, and compliance frameworks through which financial systems are governed. Regulatory sandboxes have emerged as a prominent institutional response, yet existing scholarship has not explained how they function as governance architectures or why their designs vary systematically across contexts. This study theorizes sandboxes as instruments of innovation brokerage, designed experimental regimes through which regulators convert uncertainty into governable experimentation, and identifies four brokerage mechanisms: containment, pathways, orchestration, and translation. Using directed qualitative content analysis of official regulatory documents from Kenya, Nigeria, South Africa, and Rwanda, we find two dominant configurations. Where regulatory authority is unified and procedural capacity is strong, sandbox design prioritizes bounded experimentation with predetermined constraints and formalized exit governance. Where authority is distributed and capacity is constrained, design prioritizes staged multi-regulator coordination and regulatory clarity-making. Across both configurations, exit governance, institutionalized procedures converting temporary permission into permanent regulatory classification, distinguishes disciplined experimentation from indefinite tolerance. The study advances regulatory decision science by specifying how procedural choices translate into an experimental governance architecture, extends innovation intermediation theory by showing that state actors broker through enforceable design, and provides a predictive typology that explains sandbox design variation as a systematic response to governance conditions.