Artificial Intelligence, Legal Change, and Separation of Powers

21 Pages Posted: 4 Oct 2019 Last revised: 15 Jun 2020

Date Written: September 24, 2019


A number of prominent contemporary legal scholars have recently argued in favor of replacing at least some human legal decision making with Artificial Intelligence ("AI"), assuming that AI technology improves to a level these scholars deem appropriate. This article disagrees, particularly as regards Article III judges, for two main reasons. First, human judges must strike a delicate balance between stability and change; that is, between respect for precedent on the one hand, and adapting the law to unforeseen circumstances on the other, thus playing an important role in shaping the law that is not adequately considered in this literature, and that an AI judiciary may not be able to adequately replace. Second, the loss of human judges would likely lead to a loss or diminishment of the human legal community, such that fewer people would be paying attention to the law. This community of people with strong incentives to pay attention to the law is built around the Article III judiciary, and the diffusion of knowledge throughout this community may be a significant source of the judiciary's power to fulfill its role as a check on the other two branches. The potential benefits of an automated judiciary can be better achieved in other ways, and likely do not justify the risks. At the least, these concerns are not adequately addressed by those advocating for AI judges, and should be seriously considered in the context of any effort to automate parts of the judiciary.

Keywords: Artificial Intelligence, Separation of Powers, Law & Technology

Suggested Citation

Michaels, Andrew C., Artificial Intelligence, Legal Change, and Separation of Powers (September 24, 2019). 88 University of Cincinnati Law Review 1083 (2020), Available at SSRN:

Andrew C. Michaels (Contact Author)

University of Houston Law Center ( email )

4604 Calhoun Road
Houston, TX 77204-6060
United States

Here is the Coronavirus
related research on SSRN

Paper statistics

Abstract Views
PlumX Metrics