Hallucination Rate (Legal AI)
Hallucination rate is the percentage of AI-generated legal outputs containing factual errors — including fabricated case citations, incorrect holdings, invented statutes, or misattributed legal positions — measured across a standardized test set.
Last reviewed: 2026/05/25
Definition
Why It Matters for Lawyers
How AI Tools Handle It
Frequently Asked Questions
- What is hallucination rate and why does it matter in legal AI?
- Hallucination rate is the percentage of AI outputs that contain fabricated or factually wrong content. In legal AI, hallucinations are not trivial errors — a fabricated case citation submitted in a court filing can result in sanctions, as demonstrated in Mata v. Avianca (S.D.N.Y. 2023). A lower hallucination rate means a higher proportion of AI outputs are factually reliable, reducing but not eliminating the attorney's verification burden. No legal AI tool currently achieves a zero hallucination rate.
- Which legal AI tool has the lowest hallucination rate?
- The Stanford RegLab's 2024 independent evaluation found that both Lexis+ AI and CoCounsel achieved approximately 17% hallucination rates on legal research tasks — the lowest measured among tools tested. Westlaw AI-Assisted Research measured at approximately 33%. A baseline GPT-4 prompt without legal-specific grounding measured at approximately 88%. These figures apply specifically to legal citation accuracy on the test set used; performance on other task types may differ.
- Does a 17% hallucination rate mean 17% of citations are wrong?
- Not necessarily — it means 17% of responses in the evaluated test set contained at least one hallucinated element. The nature and severity of errors varied: some were outright fabricated citations, others were misattributed holdings, and some were real cases cited for incorrect propositions. A 17% rate across responses does not mean 17 of every 100 cited cases are fabricated. But it does mean an attorney cannot skip verification — any single output may contain an error regardless of the overall rate.
Related Concepts
AI Hallucination in Legal Research
AI hallucination in legal research is when a generative AI system produces case citations, statutes, or holdings that appear authoritative but are factually false or entirely fabricated.
CapabilityCitation Validation in Legal AI
Citation validation in legal AI verifies that every case, statute, or regulation cited by an AI system actually exists, is accurately quoted, and still stands as good law — the essential check against hallucination.
CapabilityLegal AI
Legal AI refers to software systems that apply machine learning and natural language processing to automate or assist with legal tasks such as contract review, research, drafting, and compliance monitoring.
Related Tools
- CoCounsel Legal
Thomson Reuters' GPT-backed legal research and drafting with Westlaw integration (relaunched as CoCounsel Legal, 2025).
- Westlaw Precision AI
AI-powered legal research with citation-validated answers from Westlaw.
- Paxton AI
Purpose-built US legal AI covering research, drafting, and compliance.
Last reviewed: 2026/05/25. Definitions are written by the LawyerAI Editorial team. We do not accept affiliate commissions; Featured placement is clearly labeled and does not influence editorial content.