This scoping study was jointly commissioned by the evaluation units of the Adaptation Fund (AF), Climate Investment Funds (CIF), Global Environment Facility (GEF), and Green Climate Fund (GCF). The study was carried out by Elinor Bajraktari, an individual consultant. Special thanks are extended to Vladislav Arnaoudov (AFTERG), Anupam Anand (GEF IEO), Brittney Melloy (CIF) and Yeonji Kim (GCF IEU) for their guidance and supervision during the entire process. Without their assistance, this study would not have been possible.
Key insights
AI adoption in evaluations is growing but still limited, especially in climate-related work.
Strongest value: qualitative data analysis, pattern detection, and handling large datasets.
Human judgement remains essential, especially for context and causality.
AI offers major opportunities: efficiency gains, automation of repetitive tasks, and better analysis of complex climate programs.
Key risks: bias, hallucinations, weak causality, data privacy concerns, and unequal access across countries.
Responsible use requires: validation, human oversight, ethical frameworks, secure data systems, and capacity building.
Way forward: gradual adoption, customized tools, experimentation, and continuous learning.
