Artificial intelligence possibilities to improve analytical policy capacity: the case of environmental policy innovation labs and sustainable development goals

Abstract

Policy analysts dedicate a great deal of their time performing routine tasks including collecting information, identifying policy issues and options, and appraising policy options. The amount of information available online has become increasingly overwhelming. This “experience” paper examines how readily available AI tools can assist policy workers and researchers. To do so, we examine environmental policy innovation lab (EPIL) websites using three widely used generative AI programs (ChatGPT, Claude AI, and Perplexity) to assess how well they collect information about PILs and how they utilize the UN Sustainable Development Goals (SDGs). To do so, five questions are asked, including defining a PIL, whether the PILs in our database are indeed policy labs and the extent to which PILs explicitly or implicitly contribute to and/or address SDGs. Our results suggest that rapidly emerging AI tools can significantly supplement routine policy analysis and improve the routine tasks associated with analytical policy capacity. However, we conclude that despite the rapid developments, augmented intelligence is not a substitute for human analysis but rather a complementary tool.

Keywords

• Policy analysis
• artificial intelligence
• sustainable development goals
• policy innovation labs
• policy capacity