GOVERNANCE-EMBEDDED ARTIFICIAL INTELLIGENCE FOR CLIMATE-DRIVEN TRANSBOUNDARY PEST EARLY WARNING SYSTEMS

This proposal introduces Governance-by-Design Theory and its operationalization through the CAMEAL Governance Operating System (GovOS) for climate-driven transboundary pest early warning systems. Focusing on Fall Armyworm (Spodoptera frugiperda) in Kakamega County, Kenya, the research addresses critical governance deficits in existing AI-enabled agricultural systems. It prioritizes embedding accountability, accessibility, adaptive learning, and human oversight as core system functions rather than post hoc compliance measures. Seven testable hypotheses evaluate the impact of governed AI on decision quality, transparency, and institutional trust. The study employs design science research methodology and contributes 15 original frameworks, emphasizing equitable climate resilience in African contexts.

Keywords: governance-by-design, early warning systems, Fall Armyworm, climate resilience, responsible AI, accountability, accessibility, adaptive learning, human oversight, CAMEAL framework