Challenges in Reviewing Automated Decision-Making in The UK Administrative Justice

Administrative justice relies on two inseparable processes: the administrative decision-making process and the review process of these decisions through judicial review, appeals in tribunals and complaints to the ombudsman, or internal reviews. Public bodies are increasingly using AI and machine learning to automate the decision-making process, primarily to reduce costs and improve efficiency. This shift fundamentally changes the nature of administrative justice. However, numerous administrative wrongs have been encountered in automated decision-making process, such as illegality and errors. Accordingly, these wrongs in this type of decision have led to grievances and cases against the public bodies that used these AI-based systems through judicial review.
While existing literature focuses on implementations of these technologies in the decision-making process, this research focuses on investigating the challenges administrative justice institutions face while attempting to review cases about automated decisions. Using a documentary analysis method and thematic analysis of UK administrative justice institutions’ practices, the research establishes a new typology of these challenges. Specifically, it identifies three core types of challenges that hinder effective oversight: (1) Lack of transparency, (2) Regulatory gap, and (3) Expertise gap within review bodies. It concludes by proposing some solutions and recommendations to address these challenges within the administrative law domain, drawing inspiration from other laws and the experiences of other countries such as the European Union, Australia, Canada, and the USA.