Deciding with Dignity: Automated Decision‐Making, the Rule of Law, and Procedural Disrespect

Abstract This paper argues that the procedural rule of law (PROL) can provide a fruitful perspective of the dangers that artificial decision‐making (ADM) poses in the public sphere. The procedural rule of law argues that the law must reflect the dignity of the legal subject. Unlike the formal and substantive definitions of the rule of law, the procedural rule of law can identify criteria for legitimate decision‐making which can justify the continued need for human decision‐makers. These criteria are reason‐responsiveness and role‐reversibility. Taken together, they require that the decision‐maker be a thing that can engage in mutual recognition and can respond to individualized reasons. Furthermore, it can identify a particular wrong of ADM (procedural disrespect) which goes beyond the existing worries over transparency and bias. The conclusion is that the rule of law does seem to rule out widespread automation and implies the need for protected pockets of human discretion. In addition, the account presented here avoids common objections to the procedural rule of law and allows for more flexibility in procedural design than other, more well‐known accounts.