Legal AI covers systems applied to legal tasks, ranging from general-purpose models used by lawyers to purpose-built platforms trained and engineered around legal workflows, document types, and verification requirements.
The meaningful distinction is not whether a model was trained on legal text but whether the surrounding system enforces grounding, citation, permissions, and review appropriate to professional obligations.
Alternative Names:
AI for Law, Legal Artificial Intelligence
Why it Matters?
Buyers routinely conflate a general model with a legal product, and the difference shows up in the parts that are not the model: how documents are ingested and linked, whether output cites to specific pages, how matter-level permissions work, and whether outputs are validated against real case files. Those engineering choices determine whether verification is fast enough to actually happen.
Frequently Confused with
Related terms
Frequently asked questions
What makes a tool legal AI rather than general AI?
Does legal AI require a model trained on legal text?
