About The Workshop
SW 65– Artificial Intelligence and Challenges to the Rule of Law
Convenors: Pascal Felix Meier; Matthias Hächler
Contact: matthias.haechler@ius.uzh.ch ; pascal.meier@ius.uzh.ch.
Artificial intelligence is rapidly transforming the institutional landscape within which law operates. Large language models are deployed to draft legal documents, summarize case law, write law commentaries, and to assist judicial decision-making in courts. Automated systems handle administrative procedures. Predictive policing algorithms shape the exercise of public authority. These developments are not merely technical. They raise fundamental questions about the normative foundations of legal order. In particular, they raise concerns about how to safeguard the rule of law.
This workshop brings together scholars working at the intersection of legal philosophy, political theory, and the study of artificial intelligence to examine AI’s dual potential as both an opportunity and a threat to rule-of-law governance. The rule of law is a contested concept, yet its central commitments are broadly shared: that the exercise of public power should be constrained by general, prospective, and intelligible norms; that individuals should be able to orient their conduct according to law; that like cases should be treated alike; and that those subject to legal authority should have access to meaningful review and justification of decisions that affect them.
Depending on the institutional framework and the role in which AI is deployed, AI might put these commitments under pressure – or help realizing them to a fuller extent. The workshop invites contributions addressing, but not limited to, the following questions:
- Transparency and intelligibility
The rule of law demands that legal norms and the reasons for their application be accessible and comprehensible. Many AI systems operate as epistemic black boxes. Can the requirement of legal transparency be satisfied where decisions are generated or substantially influenced by systems whose internal reasoning is opaque even to their designers? Is the pursuit of “explainable AI” a genuine remedy or merely a superficial concession that leaves deeper opacity intact? Are AI decisions necessarily less opaque than those of human judges?
- Generality and the displacement of rules by patterns
Machine learning systems do not apply rules in any recognizable jurisprudential sense. Rather, they identify statistical regularities across datasets. What consequences follow for the ideal of legal generality? How should we understand the relationship between algorithmic, probabilistic “predictions” and qualitative, legal reasoning? Does AI in fact unfold the law lege artis?
- Equality, bias, justice and legal innovation
AI systems trained on historical data risk reproducing and entrenching existing patterns of discrimination. What does the rule of law require by way of safeguards against algorithmic bias, and how do these requirements relate to broader debates about formal and substantive equality before the law? Can AI, conversely, serve as an instrument for correcting human biases in legal decision-making? How can the law continue to “work itself pure” when adjudication is in the hands of algorithms that merely reproduce patterns? What ensures the alignment of AI-generated law with human justice?
- Accountability and the attribution of responsibility
Where an automated system contributes to a legally consequential decision, who bears responsibility? How do existing frameworks of legal and moral accountability adapt – or fail to adapt – to distributed human-machine decision processes? Does the insertion of AI into the chain of legal reasoning create accountability gaps that are incompatible with the rule of law? Does a “human in the loop” suffice to ensure accountability?
- Judicial authority and legitimacy
Some jurisdictions have begun experimenting with AI-assisted adjudication. What are the conditions under which such assistance remains compatible with our normative expectations for judicial decision-making and judicial independence? Is there a principled distinction between AI as a tool that supports judicial reasoning and AI as a system that supplants it? What role does the concept of human judgment play in legitimating the exercise of adjudicative power?
- Access to justice and democratic governance
AI may lower the cost of legal services and expand access to legal information. At the same time, the concentration of AI development in private hands raises questions about the democratic governance of tools that increasingly shape the administration of justice. How should institutional regulation engage with this tension?
The workshop will consist of several presentations, each followed by an open discussion. We welcome contributions from legal philosophers, political theorists, and scholars of AI ethics. Both established and early-career researchers are encouraged to submit abstracts of max. 300 words in English, German, or French to matthias.haechler@ius.uzh.ch and pascal.meier@ius.uzh.ch. The workshop aims to foster a rigorous, interdisciplinary exchange that moves beyond programmatic statements about AI and law toward a deeper examination of the normative stakes involved.
