LegiMatch vs ChatGPT for checking legal citations in a brief
By LegiMatch Research Team · Published October 17, 2026 · 4 min read
Short answer: use ChatGPT (or Claude, Gemini, Copilot) for drafting and brainstorming, not for verifying citations. A general chatbot answers from patterns in its training data; it doesn't open the court's opinion, so it can confirm a case that doesn't exist, attach the wrong name to a real citation, or "quote" language the court never wrote. LegiMatch Brief Check compares every citation in your brief with the stored text of the actual court opinion - name, pages, quotations and later treatment - and links each result to the opinion so you can confirm it.
Side by side
| General AI chatbot | LegiMatch Brief Check | |
|---|---|---|
| What it checks against | Its training data and, sometimes, web search results | The stored text of real US court opinions, with original page numbers |
| Can it confirm a case that doesn't exist? | Yes - this is how invented citations reach court filings | No - a citation that isn't in the library is reported as "not found", never guessed |
| Case name vs citation | May not notice a mismatch | Flags when the name in your brief doesn't match the case at that citation |
| Pin cites | Can't reliably tell which page a passage is on | Flags a pin cite outside the opinion's pages |
| Quotations | May reproduce or invent plausible language | Checks the quoted words against the opinion text |
| Later treatment (good law) | Unreliable, and blind to recent decisions | Flags cases treated negatively by later decisions in the library |
| Proof | A confident answer | A link to the opinion for every citation |
| Same input, same answer? | Not guaranteed | Yes |
| Table of authorities | Must be built and checked by hand | Generated with the report (PDF or Word) |
| Confidentiality | Depends on the account and settings | Uploads encrypted, stored on US servers, never used for AI training |
In our own test of five AI chatbots on 10 citations and 6 quotations, all five caught the invented cases and altered quotes, but ChatGPT and Copilot confirmed a pin cite that falls outside the opinion; LegiMatch got every item right. See the full test.
Why chatbots get citations wrong
A large language model writes the most likely next words. Case citations follow a very regular pattern - Name v. Name, volume, reporter, page, court, year - so a model can produce a citation that looks perfect and doesn't exist. Asked "is this case real?", the same model can say yes for the same reason. That is the failure behind Mata v. Avianca (S.D.N.Y. 2023) and the sanctions decisions that followed it; see AI-invented case citations: the sanctions every lawyer should know.
Web-connected chatbots are better at finding a case, but they still summarise rather than verify: they don't systematically compare your exact name, page, pin cite and quotation with the opinion.
Why checking against the opinion works
A citation is an address: volume, reporter and page point to one specific opinion. If you hold the opinion, you can test the citation exactly - is there an opinion at that address, does its name match, does the pin cite fall within its pages, are the quoted words in its text, and what have later courts said about it. That is a lookup and a comparison, not a guess, and it gives the same answer every time.
When to use each
| Task | Use |
|---|---|
| Drafting an outline, rephrasing an argument, summarising your own notes | A general chatbot (without confidential client data, unless your account protects it) |
| Finding cases with facts like yours | A legal research tool that only returns real decisions - for example LegiMatch |
| Checking every citation in a brief before filing | A citation check against the opinions - LegiMatch Brief Check or a citator |
| Final confirmation | Reading the authorities you rely on - always |
Frequently asked questions
Can ChatGPT verify case citations?
Not reliably. It can describe a case or say a citation looks correct, but it doesn't check your citation against the court's opinion, so it can confirm citations that don't exist. Verify citations against the opinion text in a legal database or with a citation-checking tool.
What is the best way to check a brief for AI-hallucinated cases?
Compare each citation with the actual opinion: confirm the case exists at that citation, the name matches, the pin cite is within the opinion's pages, quotations are word-for-word, and the case is still good law. LegiMatch Brief Check does this for every citation in a brief and links each to the opinion.
Is LegiMatch an AI tool?
LegiMatch is a legal research tool that uses AI to help find and explain relevant cases. Its answers always come from its library of real court decisions - it never invents a case, and every result links to the court's text.
Is LegiMatch cheaper than Westlaw or Lexis for checking briefs?
Brief Check is unlimited on every LegiMatch plan, from $99 a month with no annual contract, and in the 14-day free trial. LegiMatch's good-law flags are based on later decisions in its own library and are not a full replacement for KeyCite or Shepard's.
Is it safe to paste a client's brief into ChatGPT?
It depends on your account type and settings, and on your duty of confidentiality. Review the provider's data-use terms and ABA Formal Opinion 512 before sharing client information with any AI tool.
Sources
- Mata v. Avianca, Inc., No. 22-cv-1461 (PKC) (S.D.N.Y. June 22, 2023)
- Magesh, Surani, Dahl, Suzgun, Manning & Ho, "Hallucination-Free? Assessing the Reliability of Leading AI Legal Research Tools", Stanford RegLab / HAI (2024)
- ABA Standing Committee on Ethics and Professional Responsibility, Formal Opinion 512, "Generative Artificial Intelligence Tools" (July 29, 2024)
This article is general information about legal research, not legal advice. Verify every authority before relying on it.
All articlesMore from the blog
- Can AI find a case from its facts alone? A published test of similar-case searchWe rewrote 80 real court decisions in plain words, removed every name and citation, and measured how often LegiMatch Case Match finds the exact original among 4.7 million U.S. court decisions - with the method, so anyone can repeat it.
- Cases you may have missed - finding authority your brief doesn't citeAfter LegiMatch checks your brief, one click runs a fact-based search on the brief itself and lists strong matching decisions you haven't cited - how it works, what it costs and how to use the results.
- How accurate is LegiMatch Brief Check? A blind test on real court documentsOur blind test of the brief checker, with a detection rate for each kind of citation mistake - wrong court, year, pin cite, case name, quotation, invented case, wrong volume, swapped citations and statutes - plus how often it wrongly flags correct citations, measured on real court decisions and U.S. Supreme Court briefs nobody had used before.