This report attempts to take stock of predictive justice in criminal matters in France, in a context where questions and research on this subject are multiplying as fast as technologies using artificial intelligence are progressing. Without claiming to be exhaustive, it first looks at the concept of predictive justice in France, for which there is still no consensus due to the lack of a legal definition. It then identifies the practices and tools implemented in the courts, often on an experimental basis, before analysing the reception given to these technologies, both by the doctrine and by the practicioners and professionals using them. Finally, the report draws up an inventory of the regulations and other standards governing the use of artificial intelligence in the administration of justice, and highlights the various problems posed by these technologies, particularly with regard to the fundamental rights associated with criminal procedure.