France Takes Aim Against Legal Analytics

|The Volokh Conspiracy |

According to a story in the Artificial Lawyer [here], France has enacted a new law that will criminalize the "reuse" ("reutilisation") of the "identity data of magistrates and members of the judiciary"("Les données d'identité des magistrats et des membres du greffe") for the purpose of "evaluating, analysing, comparing or predicting their actual or alleged professional practices" ("pour objet ou pour effet d'évaluer, d'analyser, de comparer ou de prédire leurs pratiques professionnelles réelles ou supposées").  [The French version can be found here]

The target, apparently, are companies offering legal analytics and modeling services designed to find patterns in the decisions of particular magistrates and to predict their future decision-making performance, which have sprung up in the wake of the French government's (laudable) efforts to make all case law publicly-accessible online. [A good example of the Doctrine of Unanticipated Consequences]

I have no idea whether this law will be challenged, nor what the grounds for such a challenge might be under French law (though it is I think quite clear that such a law would be very difficult to sustain against a First Amendment challenge here in the U.S.).

Editor's Note: We invite comments and request that they be civil and on-topic. We do not moderate or assume any responsibility for comments, which are owned by the readers who post them. Comments do not represent the views of Reason.com or Reason Foundation. We reserve the right to delete any comment for any reason at any time. Report abuses.

16 responses to “France Takes Aim Against Legal Analytics

  1. Well, since modern law appears to be purely arbitrary, seems logical to shop for judges.

  2. From what little exposure to French constitutional law I’ve had, if anything they’ve got an anti-1st amendment in place.

  3. The obvious move to avoid a French law is, of course, to operate such services overseas. Probably outside the EU, too. So, Switzerland or Quebec?

    1. Why not just operate in a country with robust protections for free speech, like the U.S?

      Surely there are some enterprising French-speaking US attorneys who would be interested in taking on this project.

      1. Because US companies have such robust protection for free speech that they’re allowed to ‘voluntarily’ censor you, and will.

        Whereas some states with less free speech protection actually might prohibit that.

      2. Even among US attorneys who speak French, most are trained in US law, which is to say the tradition of Anglo-American common law. Quebec and Switzerland don’t just have a deeper pool of French-speakers, with a higher percentage of truly fluent speakers, but those lawyers are trained in French-style civil law.

  4. It seems to me it’s pretty common for legal briefs to parse things judges said in the past in order to predict how they will rule in the future, for example, this case. Doesn’t come under the statute if a judge is identified by name? There’s nothing in the quoted language that remotely suggests application would be limited to computer analytics. Ordinary human thought, the sort lawyers routinely engage in when they do legal research, is also a kind of analysis.

  5. For example, in the recent Senate confirmation hearings for Justice Kavanaugh, there was a great deal of this going on. Lots of people parsed his past statement, using his name (identity) specifically, to try to compare, analyze, etc. his actual professional practices. Indeed, until the accusations happened, that’s pretty much what the hearings were about. Not only did every participant in the process do it, so did lots of people in the media, on blogs, etc.

    1. ” Lots of people parsed his past statement, using his name (identity) specifically, to try to compare, analyze, etc. his actual professional practices. Indeed, until the accusations happened, that’s pretty much what the hearings were about.”

      And then they realized that:
      1. The Republicans had a majority in the Senate.
      2. The filibuster is dead for all judicial nominations. There was no way to stop the confirmation without convincing Senate Republicans to vote against confirmation.
      3. Those arguments were not going to stop Kavanaugh’s confirmation because everything out of that analysis that they saw as a negative, the Senate Republicans saw as a plus.

  6. I don’t see how this could be enforced. Openly selling such analyses would expose the vendors to prosecution, but since the data is publicly available, lawyers equipped with readily available software could do their own analysis and apply it, without mentioning it.

    1. Outsiders prevented, insiders not prevented

  7. France (and E.U.) has (have) a different point of view of “justice” services as U.S. do.

    For us (french), the judges are just selected as “tools”. They are only interpretating law to end a litigation. French judges are public servants. They are not supposed to have any opinion (beside the legal one) Furthermore, profiling a judge through his/her opinion is also trespassing the personal data (GDPR) prohibition.

    Therefore, there are 2 fundamental limits:
    1°) the right to be judged by a neutral judge (and not a chosen one for his/her opinion);
    2°) the right for that judge to don’t be classified for their ideas/race/sexe and so on.

    Btw, we do have (are supposed to, at least ) a robust freedom of speech / expression.

    Cheers

    1. The U.S. also believes that judges are unbiased “tools” whose job it is to apply the law and, where necessary, interpret the law.

      We also believe that judges should not be selected based on their race, sex, etc.

      But we recognize that judges do have biases. There are differences in judicial philosophy, particularly as applied to the US Constitution. We recognize (as a system, maybe not always individually) that reasonable judges can reach different results in applying the same law to the same set of facts.

      It is perfectly reasonable for a lawyer to want to know a judge’s philosophy before appearing before that judge.

    2. Re: GDPR, I believe Article 22 prohibits fully automated profiling that produces “significant effects” concerning the target. This law does not seem to have those limitations.

  8. […] Moving against emerging litigation analytics and prediction sector, France bans publication of statistical information about individual judges’ decisions on criminal penalty [Artificial Lawyer, ABA Journal, David Post] […]

Please to post comments