Key points are not available for this paper at this time.
Due to the significant advancements in technology and computing power over the past few decades, artificial intelligence-based systems have emerged as desirable tools for improving public services. Scholars have dedicated considerable effort to understanding this growing trend. A key focus of this research has been accountability, which has become a prominent concern in the scholarly discourse. Studies reveal that the integration of AI into government processes introduces unique challenges to maintaining governmental accountability. In response, scholars have proposed various policy and management strategies to tackle these issues. Despite extensive research, there remains a divergence in how accountability is conceptualized in the context of using AI-based systems in government, indicating a gap in the systematic understanding regarding how to hold the use of such systems accountable. To bridge this gap, we conducted a systematic review of the existing research on AI-based system accountability in the public sector. Our findings based on thirty-six articles highlight the existence of different forums and actors, and the mechanisms recommended to ensure accountability. We also identified challenges and enablers that could affect the efficacy of these mechanisms. Lastly, our analysis uncovers several areas requiring further exploration, offering directions for future research in this field.
Yuan et al. (Sun,) studied this question.