The Great Disruption and “Computational Jurisprudence”

Last May John O. McGinnis and Russell G. Pearce wrote about the “great disruption” in an article published in the Fordham Law Review. They began by stating that, “Law is an information technology–a code that regulates social life.” They concluded that “the disruptive effect of machine intelligence will trigger the end of lawyers’ monopoly and provide a benefit to society and clients as legal services become more transparent and affordable to consumers, and access to justice thereby becomes more widely available.” They also noted that,

The market for electronic legal services is at a relatively early, yet significant, stage in terms of the disruptive effect of machine intelligence in undermining lawyers’ monopoly. As machine intelligence in lawyering develops exponentially, it will take an increasingly larger role in five areas of legal practice: discovery, legal search, generation of documents, creation of briefs and memoranda, and predictive analytics. Eventually, machine intelligence will prove faster and more efficient than many lawyers in providing those services.”

McGinnis spoke about these ideas a couple of weeks ago at the first CodeX Speaker Series event held at the Stanford Law School this year. He also explored changes in legal search, the increasing capability of machine intelligence and its impact on jurisprudence.

If we think that law is itself a kind of information technology, which it ultimately is, it informs people of the standards that we expect from them and tries to take information from the world, it shouldn’t surprise us at all that a revolution in information technology, which is what the computer is, is also necessarily a revolution not only in the legal profession but in legal jurisprudence.”


  1. It is interesting to think about technological developments alongside the ABS discussion.

    What sort of responsibility does a search engine designer owe to a client?

  2. There were a couple of comments during the Q&A that touched on some of the potential economic aspects of these developments. Using machine intelligence to build new economic models, for example. But the point was also made that data alone is not enough, hypotheses and experience are needed to develop these models. There will be competition in this area.

    It was suggested that creating models without industry knowledge make building these things difficult and lawyers might consider learning new skills, like math, so they can develop these models from the inside. Which I think is beginning to happen to some degree.

    McGinnis then raised the impact this might have on the future of legal education. He asks: Where is the best place to train people? And how do we marry data analysis with the necessary knowledge?

    Business students are traditionally more enthusiastic about math and data and he suggests that “law schools need to be a little more visionary.” Add statistics and data analysis to the JD education. Or perhaps a Master of Legal Science for software engineers to create opportunities for cross-pollination in the law school environment.

  3. Tim — The potential of legal information technology raises a whole series of questions. For example:

    1) Is it really shortsightedness or a failing of the law school curriculum not to offer courses in “math” or data analysis? Is it even in the province of law schools to teach legal technology, and could they teach it sufficiently to the task? Wouldn’t it make more sense for a Master of Legal Science program to be offered by a library school (school of information studies)? Or in a computer science department? Is a “law degree” essential to the task? I don’t know what degree qualifications the programmers building IBM’s Watson have, but they’ve been doing pretty well at winning Jeopardy games. Would they really need a law degree to scale Watson for law? Because we’re all looking at the problem from within the legal profession, we tend to assume the solutions will (must?) come from within the legal profession and require legal credentials. (It’s an extension of the old question: Do you need a law degree to be a good or effective law librarian?)

    2) Most likely, the new legal information technologies will be developed within the existing legal information or business information industries. Instead of “legal research according to West”, we could well end up with “legal research by West”. As Michael asks above, I wonder what kind of responsibility their shareholders will think they owe to a client? Who is their client? Is it the the lawyer (or other legal or para-legal professional) using their service? Or is it the legal services consumer at the further remove? Certainly, the information industries have no such responsibilities in the existing culture.

    3) With the legal research function in the hands of the commercial sector, what is the real likelihood of legal services becoming more affordable and more accessible?

    Legal information technology has the potential to make many things better, but also worse.

  4. I think you’ve captured the concerns of some of the participants in the room. Namely that lawyers/law students are not usually concerned with statistics, mathematics and data analytics. And for sure, maybe they shouldn’t be. But some also seemed to feel that if lawyers don’t acquire these skills then others may develop and offer legal services previously provided exclusively by lawyers. And maybe that’s not a bad thing. So not a failing per se but perhaps an opportunity for law schools to offer something like the Katz and Bommarito course on legal analytics.

    The IBM developers have been consulting with domain experts to assist with their various flavours of Watson. So the programmers may not be legal experts but they are working with and receiving advice from those with experience in the field.

    Many of these developments are driven by consumer resistance to the perceived high cost of legal services. And if they can get improved/cost effective results from a lawyer using legal analytic tools they’ll probably use that service. As McGinnis says, “it’s the combination of a machine and a lawyer that will be better, not that the lawyer will be replaced.”