Artificial Intelligence and Administrative Discretion in Disciplinary Procedures: Between Technological Efficiency and Legal Limits
We are living through a new Industrial Revolution — a new era in the domain of administrative proceduralisation. We are witnessing a digital transition in which administrative procedures are already being conducted in an automated manner, fundamentally altering the typical paradigm of legal-administrative relations. This relationship can no longer be conceived as involving only two parties, as we are now faced with the presence of a “third” actor that transforms the entire decision-making dynamic of the Administration: artificial intelligence. There are administrative procedures that are fully processed in an automated manner, with the corresponding decision being rendered by an algorithm through an electronic administrative act. Now, considering that the disciplinary procedure is one which may lead to the adoption of an administrative act of a punitive nature — often restricting rights, freedoms, and guarantees — and that the legislator has established enhanced defence safeguards in this context, it becomes crucial to examine the potential suitability of automating disciplinary procedures and the possible replacement of the instructing authority by an algorithm. Moreover, in the realm of disciplinary procedures, there are moments of administrative discretion, which makes it equally important to assess the feasibility of using automation in the context of discretionary powers, exploring its application across the various forms in which such powers are attributed. To this end, we will analyse the traces of discretion present in the different phases of the disciplinary procedure, evaluating the possibility of automation in this domain and the viability of producing an automated administrative act as a result of the exercise of discretion.
Publication Details
- Journal
- Scientific Repository of Open Access of Portugal (RCAAP)
- Published
- 2025-07-18
- Primary Topic
- Ethics and Social Impacts of AI
- Type
- article
- Field-Weighted Citation Impact
- 0.00