• Andrew C. Michaels

Abstract Innovation, Virtual Ideas, and Artificial Legal Thought


Questioning Artificial Intelligence in Law: Is it a good idea to replace legal thought with artificial legal thought? Even assuming, arguendo, that doing so would enable more consistent and efficient application of the law as it stands, I’m skeptical.

Our current legal system could in fact be seen as quite inefficient. For example, we have hundreds of judges across the country, applying the law in (ideally) more or less the same way. There is redundancy here, but there may be real value in the redundancy, in that the system we have now fosters a large community of people with a strong incentive to pay attention to the law. Cf. John M. Golden, Redundancy: When Law Repeats Itself, 94 Tex. L. Rev. 629, 629 (2016) (“The pervasiveness of legal redundancy has at least one straightforward explanation. Redundancy has much to offer.”).

In our current system, even those outside of this community have some incentive to think about the law. Ordinary people, acting through judges and lawyers as intermediaries, can potentially change the law through argument. It is not easy, but it does happen. Litigation may drag out for years, through multiple appeals, involving intense adversarial argumentation and serious consideration from multiple layers of judges, (not to mention their law clerks). But this seemingly inefficient process allows for human deliberative thought, which may be exactly what we need when considering whether to alter the law. A decision is then rendered explaining the outcome, again connecting law with the public.

Now imagine a future where much of this has been replaced by some sort of centralized Artificial Intelligence system. Citizens wear a communications device that instructs them “exactly how to comply” with the law in any given situation, such that all they have to do is “obey simple directives.” See Anthony J. Casey & Anthony Niblett, The Death of Rules and Standards, 92 Ind. L. J. 1401, 1404, 1435 (2017). When the only rule that the citizen needs to know is: obey the machine, the law becomes a pure assertion of authority. This future implies at least a “reduced role for judges,” id. at 1405, and thus seemingly for lawyers as well. Society becomes free of any need to think about the law, when the law machines do our thinking for us.

Such a future would probably be more efficient, but that doesn’t ipso facto make it more desirable. Cf. Oil States Energy Servs., LLC v. Greene’s Energy Grp., LLC, 138 S. Ct. 1365, 1380 (2018) (Gorsuch, J., dissenting) (“A judicial hearing before a property interest is stripped away . . . can slow things down. But economy supplies no license for ignoring these – often vitally inefficient – protections”) (emphasis added). Once we have lost the incentive to pay attention to and question the law, we have likely lost the sense that the law is something we as a society create to govern ourselves. We would also seem to have lost, to a large degree, the possibility of effecting change in the law through reasoned, deliberative, public argumentation. And in such a future, if and when the law does change – considered and deliberate legal change being a being an important feature of our current legal system – it’s not at all clear who would do the changing, for what purpose, or who would notice.

As I discuss in my recent article, Abstract Innovation, Virtual Ideas, and Artificial Legal Thought, in the Journal of Business & Technology Law, before we blindly rush head first into a future of automated law, it is at least worth thinking about these potential issues now, while we have the capability and some incentive to do so.

Andrew C. Michaels, University of Houston Law Center


Recent Posts

See All

Copyright © 2018 All Rights Reserved. Faculty of Law, The Chinese University of Hong Kong

The Chinese University of Hong Kong