Archives quotidiennes : octobre 23, 2016

New technologies and legal practice – a French professor’s perspective (1/4)

NB: this is an English expanded version of a previous post.

The sector of legal practice is teeming with technological innovations. What we call “the Uberization of lawyers” is on. Everyday we read about a new app, a new algorithm, a new way to enhance our activity as legal practitioners or researchers. At the same time it is disturbing to hear about the coming changes, and exciting for lawyers and all those that evolve in the legal sphere to think about the new possibilities, the new opportunities offered by the use of technology. This affects both practice and teaching of law. For instance, the MOOCs can be used for basic training in law but also to update a practicing lawyer’s skills, or to pass legal information to his/her clients.

There is for the lawyers a key issue in any new application of technology to the legal activities. That risk is to loose sight of the contribution of lawyers, of the value they add, of the knowledge they convey. If new technologies allow legal information to flow more freely than ever, are lawyers still useful?

Cynics will say: have lawyers ever been useful? Couldn’t we say that the lawyers’ role is to create problems, where there would otherwise only be interesting projects? The geek who has a great idea for a new application for iPhone and questions a lawyer may hear some refraining words: the lawyer will talk to him about third parties’ rights, the need to protect the geek’s creation, and so on. In short, the lawyer will remind his/her creative-minded friend that the law is made of rules, and that they are, by definition, binding … But lawyers have to assume that. Our society cannot function if it is not regulated by the law. Therefore, it is necessary to have a category of people, the lawyers, dedicating their lives to understand the law and explain to others how to move through the legal system. Yes, the law could be simpler, could change less often, but that’s another story.


Let’s return to new technologies. In recent months, several earth-shattering announcements were made about the application of artificial intelligence (AI) to the field of legal practice. Maybe the most striking announcement was about BakerHostetler, a US law firm, « recruiting » ROSS as a new lawyer. The thing is: ROSS isn’t exactly a lawyer (I’m not sure it’s even related to the Suits TV show!). Actually, ROSS is no lawyer at all. It is only a legal software.

The announcement is repeated again and again: the robot is going to be recruited by a law firm. It would have been less selling to merely tell that the firm was going to acquire the rights to use a new efficient legal software. But the word « recruit » appears quite pertinent when it is told at the same time that ROSS will replace in the BakerHostetler firm not less than fifty lawyers, which have therefore to find a new job. If ROSS is not « recruited » in a strict labour law meaning, jobs are all the same suppressed as a consequence of its integration in the BakerHostetler firm.

But what is ROSS exactly?

We understand from reading the website that ROSS is a legal data processing software, which answers legal questions directed to it in everyday language. The site takes the example of the question: “Can a bankrupt company still engage in an activity?”, to which it should be able to answer. ROSS also has other functions such as legal monitoring: if new court decisions are affecting the branch of law that interests me, then I will be informed by ROSS. But there is the hope that ROSS will operate in a selective way, in other words, in a intelligent fashion, instead of burrowing its users with tons of raw legal information.

In addition, ROSS, which is based on the IBM Watson technology is expected to become increasingly efficient as it will be used for legal research. In short, while giving the requested answers, ROSS will learn and become more and more proficient. Such a use of a legal software leads to the question of how far human intervention is replaceable. In the field of insolvency law, for instance, how many of the following tasks can be “machine-sourced”: ensuring that the creditor’s rights are taken into account by the receiver, that a collateral remains effective, negotiating with the debtor and the receiver, ensuring that the scheme of arrangement is satisfactory, that alternative solutions are provided for? If the authorities supervising the insolvency proceedings (in France, mainly the commercial court and various insolvency authorities: receiver, insolvency judge, etc.) were themselves automated and paperless entity, maybe that would speed up the replacement of the human intervention?

In the future, will the various creditors of a bankrupt company turn to the Android Sophia to negociate payment extensions or discounts, in order to save the company while ensuring the creditors get partially paid ?

Coming next: Peter


1 commentaire

Classé dans French Law, Idée nouvelle, Nouvelles pratiques, Nouvelles pratiques du droit, Uncategorized