Foundation model
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 popularized by Stanford HAI in 2021. It marks a paradigm shift: a single versatile base instead of one model per task.
Foundation models mark a turning point in the history of artificial intelligence. Before they emerged, every application required a dedicated model trained on task-specific data. Today, a single base model — GPT-4, Claude, Mistral, Llama — can be adapted to dozens of different tasks: text generation, sentiment analysis, classification, translation, automatic summarization.
The term was formalized in 2021 by Stanford HAI (Human-Centered Artificial Intelligence) to describe these models trained through self-supervised learning on massive, unlabeled corpora. The principle is simple: the model learns to predict what comes next in a text from billions of documents, giving it a general understanding of language that applies across many domains.
In the French legal sector, legaltech solutions such as Doctrine, Predictice, Jimini, GenIA-L and Lexis+AI adapt these foundation models to French law. This adaptation relies on fine-tuning on legal corpora and/or RAG architectures that ground answers in verified sources. The challenge is to turn a general language ability into reliable legal expertise.
Related terms
An LLM (Large Language Model) is an artificial intelligence model trained on massive volumes of text to learn the statistical stru…
GPT (Generative Pre-trained Transformer) is a family of AI models developed by OpenAI, built on the Transformer architecture intro…
Claude is an AI model developed by Anthropic, a company founded in 2021 by former OpenAI researchers. It stands out for its "Const…
Fine-tuning is additional training of a pre-trained foundation model on specialized data to adapt its performance to a specific do…
RAG (Retrieval Augmented Generation) is a technique that combines document retrieval with an LLM. Before generating an answer, the…