Artificial intelligence as an aid to constitutional checks on the administrative state

This contribution asks whether and how artificial intelligence (AI) could be used to improve constitutional checks on the actions of administrative actors? Failures to keep tabs on the contemporary administrative state are often linked to informational problems such as lack of transparency, information ‘overkill’ and ‘not knowing what one does not know’. One way in which AI could aid human efforts to control administrative actors is to let algorithms work alongside human-led administrative decision-making to reveal potential biases in the latter (micro context). Another is using AI-fed monitoring systems to signal certain trends within administrative agencies or in the field they administer (macro context). The contribution investigates the potential of AI applications to fulfill these legitimacy-enhancing roles in constitutional oversight of the administrative state, paying attention to the normative frameworks that could ensure that constitutional actors meaningfully stay in charge.