NLP (Natural Language Processing) is the branch of artificial intelligence dedicated to the interaction between machines and human language. It brings together a set of techniques: tokenization (splitting text into units), syntactic parsing, named entity recognition (identifying names, dates, amounts), sentiment analysis and automatic summarization. Each of these building blocks has a direct application in law.
In the legal sector, NLP is everywhere. It powers the automatic extraction of clauses from contracts, the classification of legal documents by type and subject matter, the identification of parties and amounts in court rulings, and the summarization of lengthy documents. As early as 2016, researchers showed that an NLP model could predict European Court of Human Rights decisions with 79% accuracy.
Recent advances in LLMs have dramatically expanded the capabilities of legal NLP. Today, legaltech tools combine classic NLP techniques (extraction, classification) with the generative power of large language models to deliver end-to-end solutions: from semantic search to assisted drafting and the predictive analysis of litigation.