87

This article aims to discuss the stages of the commercialization and algorithmization of justice in France. In the introduction, it questions the adaptation of justice to unintended innovations, whether financial or technical. Firstly, it addresses the commercialization of civil justice, observing the commercialization of access to justice rather than the financialization of disputes by third parties. This commercialization of access to justice is noted through the mandatory appointment of a lawyer before the judicial tribunal, mandated by the December 2019 reform of the civil procedure code, the growing role of legal protection insurance as a prerequisite for legal aid requests, the role of legal aid in financing litigants’ disputes, legaltechs offering court referral services, and the new paid access to commercial justice introduced by the law of November 20, 2023. It also notes the contractualization of criminal justice with the appearance of plea bargaining, allowing an offender to negotiate their sentence with the Public Prosecutor, and the introduction of the judicial public interest agreement, a form of conditional dismissal that has seen its scope expand. Secondly, recognizing the potential utility of AI in easing court congestion, it analyzes various uses of AI in justice, highlighting open data for court decisions, providing access to all court decisions, the creation of a neural translation aid tool for translators assisting courts, and a tool to help direct extensive briefs to the appropriate chambers in the Court of Cassation. The article concludes by questioning the role of the civil judge. Should the judge not adopt a modular procedural approach, offering the most suitable solution for the case? Should the judge not act as a guide, proposing all possible procedural alternatives for the litigants’ dispute: a judgment, a conciliation with the judge or a conciliator, mediation, arbitration, a participatory procedure, a bifurcation, etc.? The judge’s role is to direct them towards the best procedural solution to better meet their needs and demands for justice. In light of this necessary approach, the question arises about the impact of algorithmization on the daily work of judges. One could imagine an ecosystem based on the open data of court decisions that would be natively formalized. An AI could enable the modeling of decisions. This modeling does not challenge the judicial sovereignty of the judge but allows them to free up more time. Built on clear law and language, this modeling would achieve two fundamental objectives: a simple and immediate understanding of decisions by litigants and easier execution of decisions in digital form.

Cet article se veut un propos d’étape sur la question de la marchandisation et de l’algorithmisation de la justice. Il aborde dans un premier temps la question de l’accès à la justice civile en constatant une marchandisation de celui-ci par la constitution obligatoire d’avocat devant le tribunal judiciaire, le rôle croissant des assurances de protection juridique comme recours préalable à la demande d’aide juridictionnelle, le rôle de l’aide juridictionnelle, les legaltechs et le nouvel accès payant à la justice commerciale. Il constate également une contractualisation de la justice pénale avec la comparution sur reconnaissance préalable de culpabilité et la convention judiciaire d’intérêt public. Dans un second temps, il analyse les différents cas d’utilisation de l’IA dans la justice en mettant en exergue l’open data des décisions de justice, la création d’un outil d’aide à la traduction neuronale et l’outil d’aide à l’orientation des mémoires ampliatifs devant la Cour de cassation. Il se conclut en postulant d’une nécessaire approche modulaire des contentieux grâce à l’IA pour répondre au mieux aux besoins et demandes de justice.