Citation Safe vs. Lexis Protégé

Protégé is positioned as a step beyond a simple research chatbot — able to carry out multi-step tasks with less back-and-forth prompting, closer to an AI paralegal than a search assistant. That is a genuinely different product ambition from Citation Safe, which does exactly one narrow thing: checking whether citations already in a document exist, are quoted accurately, and support the proposition they're attached to.

Our separate Lexis+ AI comparison page cites an independent Stanford study measuring a hallucination rate for that specific product. Protégé is a distinct feature layer released later, and we are not aware of an equivalent independent study measuring Protégé specifically — attributing the same figure to a different product would be exactly the kind of unsupported claim we try not to make. If independent research on Protégé's accuracy is published, we'll cite it directly rather than borrow a number from an adjacent product.

Protégé is delivered within Lexis+ AI subscription tiers rather than priced as a separate line item, so its effective cost depends on which Lexis+ AI plan a firm is already on. Citation Safe's $12-$199/mo tiers are published, standalone, and don't require an existing Lexis subscription.

A firm using Protégé for multi-step drafting and research automation is still producing documents whose citations benefit from an independent, deterministic verification pass — agentic capability doesn't remove the underlying hallucination risk documented broadly in AI legal research, it just moves where in the workflow that risk shows up.

Lexis Protégé is an agentic AI layer for multi-step legal work bundled into Lexis+ AI; Citation Safe is a narrow, standalone, published-price citation verifier that checks what a workflow like Protégé's ultimately produces.