Haim V. Levy (HVL Legal & Life Sciences Advisory) has posted Machines That Reason, Institutions That Judge: Artificial Intelligence and the Nature of Judgment on SSRN. Here is the abstract:
This Article develops a jurisprudential account of judgment in light of artificial intelligence systems capable of simulating legal reasoning. Large language models can generate outputs that replicate the structure of adjudicative justification, including rule articulation, fact application, and reasoned conclusions. This capacity raises a foundational question: if the form of judicial reasoning can be reproduced without human deliberation, what constitutes judgment as a legal and institutional act? Drawing on a structured analysis of model-generated opinions across federal cases, the Article shows that justificatory form can be simulated with considerable fidelity. It argues, however, that judgment is not reducible to doctrinal accuracy or argumentative coherence. Rather, judgment is an institutional practice grounded in authority, accountability, and the public attribution of responsibility. The central implication is that while machines may generate reasons, judgment—understood as an act of legal authority—remains irreducibly institutional.
Recommended!
To receive new posts from Legal Theory Blog by email, get a free subscription to Legal Theory Stack.
Lawrence Solum