The Legaltech & Legal AI glossary

An LLM (Large Language Model) is an artificial intelligence model trained on massive volumes of text to learn the statistical structures of human language. It works by predicting t…

GPT (Generative Pre-trained Transformer) is a family of AI models developed by OpenAI, built on the Transformer architecture introduced in 2017. These generative models are pre-tra…

Claude is an AI model developed by Anthropic, a company founded in 2021 by former OpenAI researchers. It stands out for its "Constitutional AI" approach aligned with ethical princi…

A foundation model is a large AI model trained on massive unlabeled data through self-supervised learning, adaptable to a wide range of tasks without full retraining. The term was …

RAG (Retrieval Augmented Generation) is a technique that combines document retrieval with an LLM. Before generating an answer, the model retrieves the relevant information from an …

Prompt engineering is the art of crafting precise, structured instructions to get optimal results from a generative AI. It is not programming but a methodical way of formulating re…

AI hallucinations refer to the generation of false, fabricated or inconsistent information presented with apparent certainty by a language model. The phenomenon is inherent to how …

Fine-tuning is additional training of a pre-trained foundation model on specialized data to adapt its performance to a specific domain. Unlike the initial pre-training (which costs…

NLP (Natural Language Processing) is the field of AI that enables machines to understand, interpret and generate human language. It covers tokenization, syntactic parsing, named en…

Tokenization is the process of breaking text down into "tokens", the basic units processed by an AI. A token can be a whole word, part of a word, or a punctuation mark. Models have…

Embeddings are numerical representations of words, sentences or documents, expressed as vectors in a high-dimensional mathematical space (768 to 8,192 dimensions). Semantically sim…

Generative AI creates new content (text, images, code), while predictive AI analyzes historical data to anticipate future outcomes. In law, predictive AI is used by Predictice to e…

Temperature is a hyperparameter that controls how much randomness goes into an LLM's responses. A low value (close to 0) produces deterministic, predictable answers. A high value (…

Predictive justice refers to the use of AI algorithms to analyze past court decisions in order to anticipate the likely outcome of a dispute, estimate damages, and assess the chanc…

The AI Act is European Regulation 2024/1689, the world's first comprehensive legal framework governing artificial intelligence. Adopted on 13 June 2024, it follows a risk-based app…

AI ethics in law refers to the set of professional principles governing the use of AI by legal professionals: algorithmic transparency, bias prevention, confidentiality, human over…

An AI agent is a new generation of autonomous AI system able to perceive its environment, set goals, plan actions and act independently to complete complex tasks end to end. Unlike…

MCP (Model Context Protocol) is an open protocol created by Anthropic in November 2024 that standardizes how AI connects to external data sources, tools, and services. Built on JSO…

AI-powered legal automation refers to using artificial intelligence technologies to automate repetitive, low-value tasks: reading and sorting documents, extracting data, generating…

AI-assisted legal research uses semantic search and NLP to transform how legal databases are queried. Unlike keyword search, semantic search understands user intent and conceptual …

A training corpus refers to the body of data used to train an AI model. For LLMs, it consists of billions of texts drawn from the internet, books and articles. The copyright questi…

In AI, open-source models (Llama, Mistral, DeepSeek) make their code and weights publicly available and can be hosted on-premises. Proprietary models (GPT, Claude, Gemini) are clos…

Data confidentiality and AI covers the legal, technical and ethical challenges of protecting data and attorney-client privilege (article 66-5 of the French law of 31 December 1971)…

Legaltech is a contraction of "Legal" and "Technology". It refers to the full range of companies and technology solutions applied to the law. In France alone, around 164 active pla…

Legal SaaS refers to cloud-hosted software, accessed by subscription over the internet and built for legal professionals. These solutions centralize matter management, billing, tim…

No-code and low-code are platforms that let you build applications through visual interfaces, with no coding at all (no-code) or with minimal programming (low-code). For lawyers, t…

Vibe coding is an emerging approach popularized by Andrej Karpathy (OpenAI co-founder) in February 2025, in which the user describes what they want to build in plain language and t…

Legal Design is an interdisciplinary approach that applies Design Thinking methods to law in order to turn complex legal texts into clear, visual and engaging documents. Coined in …

Smart contracts are self-executing computer protocols deployed on a blockchain that automatically enforce contractual obligations when predefined conditions are met. Operating on "…

Blockchain is a decentralized, transparent and secure technology for storing and transmitting information. Applied to law, it makes it possible to certify the existence, integrity …

An API (Application Programming Interface) is a set of protocols that lets two pieces of software communicate and exchange data automatically, acting like a "digital bridge". In le…

LegalOps (Legal Operations) refers to the strategic, optimized management of an in-house legal department's operations, covering organization, digitalization, process optimization …

CLM (Contract Lifecycle Management) refers to the set of processes and tools used to manage the entire lifecycle of contracts: drafting, negotiation, approval, signature, execution…

RegTech (Regulatory Technology) refers to the technology solutions that automate regulatory compliance, primarily in the finance, banking and insurance sectors. Born in the afterma…

GovTech refers to the technology solutions designed to modernize public services. According to Roland Berger, it covers "any technology startup with a public entity as a client". T…

Legal Analytics refers to applying data analysis, statistics and AI to legal information in order to surface trends, predict outcomes and sharpen litigation strategy. The disciplin…

E-discovery (electronic discovery) is the process of identifying, collecting, processing and producing electronic documents (emails, databases, instant messaging) in the context of…

An electronic signature is a technical process for expressing consent on a digital document. The eIDAS Regulation (910/2014) defines three levels: simple, advanced and qualified (t…

ODR (Online Dispute Resolution) refers to out-of-court dispute resolution mechanisms (mediation, conciliation, arbitration, negotiation) delivered through internet platforms. The E…

A virtual data room (VDR) is a highly secure online storage space for sharing confidential documents during complex transactions (mergers and acquisitions, fundraising, audits, lit…

A Legal Prompt Library is a curated collection of natural-language instructions designed to guide legal AI tools. According to the ABA (American Bar Association), it is "the natura…

Legal open data refers to making public legal data freely available and reusable: court decisions, legislation and regulations, parliamentary records. In France, the Digital Republ…

Interoperability of legal tools refers to the ability of different legal software systems to communicate, exchange data and work together seamlessly. In law firms and legal departm…

Legal SEO (Search Engine Optimization) refers to all the organic-search optimization techniques applied to law firm websites. It rests on three pillars: content (high-quality legal…

A lawyer's online reputation is the digital image of a lawyer as shaped by search results, online reviews, social media and legal forums. In France it is governed by Article 2.2 of…

Legal content marketing is a strategy built around creating high-quality content (blog posts, guides, podcasts, videos, newsletters, webinars) to demonstrate a lawyer's expertise a…

AEO (Answer Engine Optimization) is the practice of optimizing content so it gets selected as the answer by AI-powered search engines (Google AI Overviews, ChatGPT Search, Perplexi…

GEO (Generative Engine Optimization) is an emerging discipline theorized by researchers from Princeton and Georgia Tech (2023) that aims to optimize content so it gets cited and re…

The legal SERP (Search Engine Results Page) is the Google results page for law-related queries. It includes several elements: Google Ads, the Local Pack (top 3 Maps listings), feat…

Lawyer personal branding refers to the techniques used to strategically promote a lawyer's image, expertise and values as a personal brand, primarily through LinkedIn, public speak…