Glossary
Foundation Model
A large machine-learning model trained on broad data that can be adapted to many downstream tasks.
Last reviewed 2026-08-22
In plain language
The term was popularized in academic and policy writing to describe highly general models, especially large language and multimodal models. Related legal terms include general-purpose AI models in EU law. Foundation models raise issues of evaluation, copyright, compute concentration, systemic risk and downstream accountability when many applications depend on a few providers.
Why it matters for AI governance
Governance that only regulates end applications can miss concentrated infrastructure. Conversely, regulating models without use-context can miss domain-specific harms.