Earlier this month, France passed legislation effectively banning the use of predictive litigation AI. The new law is located in Article 33 of the Justice Reform Act and aims to curb the commercialization of judicial decision-making data.
A prominent line in Article 33 states: ”The identity data of magistrates and members of the judiciary cannot be reused with the purpose or effect of evaluating, analysing, comparing or predicting their actual or alleged professional practices.”
French AI Ban. Protecting universal justice?
This targeted ban on AI resulted in part from the tremendous success of legal ventures like Ravel Law who are now able to model the decision-making tendencies of judicial decision makers with relative ease, apparently rubbing French magistrates the wrong way.
Although there’s no general consensus as to why the French ban on AI was included in the bill, many reasons have been proposed, ranging from the need for judicial anonymity, to the fear that predictive analytics might expose glaring dichotomies between objective judicial norms and reality.
Reasoning aside, Article 33 is now perhaps the strictest ban on the use of legal AI in any jurisdiction. French court decisions will still be published in full without a judges’s name redacted. However, the law bars firms from engaging in a full review of judicial rulings and carries a five-year prison sentence for offenders.
Step Forward or Step Back for Justice?
Governments around the world have a tremendous opportunity to drastically improve the objectivity of their courts by employing (or, at the bare minimum, allowing) artificial intelligence to run predictive analysis on their judges.
Daniel L. Chun, a researcher at both the Toulouse School of Economics and University of Toulouse Faculty of Law recently published a paper showing AI can accurately predict the outcome of an asylum case with over 80% accuracy (before the case opens) by knowing only the identity of the judge and the nationality of the asylum seeker. Chun attributes the tremendous predictability of asylum courts to “snap judgments” that judges make consistently over the course of their careers. Using AI, we can now highlight judges who are statistically more predictable than their counterparts, implying that some rely on “snap judgments” as opposed to a careful and un-biased deliberation of the case at hand.
AI therefore provides an opportunity to fundamentally improve our courts, ensuring judges are held accountable for biases and snap decisions. Judges are human beings after all, so what can we do about it? Well, Professor Chun speculates that by informing someone of their bias, there is often a significant improvement in the decision-making process related to the bias.
The end result? A better and more fair judicial system. The French judges that rallied to outlaw this technology have good reason to be worried. Since publication of French court decisions, there have been studies confirming the bias of French judges. AI is now blowing the whistle, alerting us that unbiased judging is not as universal as we once believed.
That does not mean we should punish the messenger. We should actively encourage technological advances that help to ensure a more fair and transparent judicial process for all.
Ian Connett, Esq. (@QuantumJurist) is the Founder of QuantumJurist, Inc., a LegalTech consulting and technology venture dedicated to improving and creating efficiencies in the legal services industry. Ian is also a Contributing Editor to the EvolveTheLaw.com Legal Innovation Center and Host of the Evolve the Law Podcast. Ian resides in New York, where he has served as an in-house counsel to numerous technology companies. You can connect with Ian on Twitter and LinkedIn and you can reach him by email at ian@quantumjurist.com (for story ideas, personal correspondence, media inquires or speaking engagements).
