Key points are not available for this paper at this time.
Globally, governments have increasingly implemented Artificial Intelligence (AI) in public service delivery and decision-making to replace human officials in the name of improving scalability, cost-effectiveness, and efficiency. However, few empirical studies have explored the challenges citizens face in seeking accountability when government AI agents fail. To fill this gap, this paper investigates how citizens perceive and attribute blame for AI-induced public service failures compared to those caused by human officials, addressing the potential ‘accountability deficit’ in AI governance. Using Weiner’s Attribution Theory as the framework, we conducted three scenario-based experiments with 516 participants. The results revealed that citizens generally blamed AI agents less than government departments due to lower perceptions of controllability over the service task compared to the same failures caused by human officials. However, when AI was identified as outsourced, their blame toward the government was significantly lower. Thus, our findings support the idea that the application of AI in public services introduces uncertainty into governmental reputation and accountability. Overall, this study contributes to the growing body of knowledge on AI in public services by underscoring the importance of developing ethical and legal governance frameworks to address potential accountability deficiencies and blame avoidance in public services using AI.
Liang et al. (Wed,) studied this question.