AI Literacy, Decision-Making Experience and Trust in AI-Generated Policies

• In Event: Citizen's Perceptions of AI

Abstract

n recent years, the widespread application of artificial intelligence (AI) has propelled AI-based decision-making into the forefront of the public sector. However, the pivotal question of policymakers' trust in AI-generated policies and the influential factors affecting this trust remains largely unexplored. This study posits that individuals’ AI literacy, decision-making experience, and specific attributes of AI tools significantly influence trust in AI-generated public policies. Through two lab experiments involving 100 public policy students and practitioners, we meticulously evaluate their confidence in AI-forged birth and ecology policies. Our initial analyses illuminate that individuals possessing real-life decision-making experience exhibit a lower propensity to trust AI-generated policies in both domains, thus exhibiting diminished inclinations to embrace AI in future decision-making endeavors. Conversely, higher levels of AI literacy correspond to a greater level of trust in AI-forged policies, coupled with an increased propensity to adopt them in future practices. In addition, compared to policies emanating from AI tools of global repute, individuals exhibit greater trust in policies generated by AI tools intimately linked with their own cultural framework. These discoveries highlight the imperative of assimilating professional experience and cultural factors to advance AI-generated policymaking, yielding critical insights into integrating AI technologies in policy formulation.

Authors

• Ruoting ZHENG, Xiamen University

• Fang Chen, Xiamen University