What is CaseRead?
CaseRead is an AI legal research service at caseread.ai. A user asks a legal question, picks federal or state jurisdictions, and gets an answer with links to the sources it found. It can also search the files you upload to a project together with public law, and it can draft documents.
The vendor describes it as AI legal research for legal professionals and says it makes mistakes, so every citation should be checked. Its pages make clear that it is not a law firm and does not give legal advice. This guide read the vendor's pages and terms on November 4, 2026.
What it does
- Research: cited answers from federal and state law, with a deep research mode on higher plans that adds a web pass.
- Hallucination Shield: you paste a brief or upload a file, and each citation is looked up, then read against the sentence it supports. The vendor says the check covers up to 50 citations per run and calls it an automated screen, not a citator.
- Projects: the parties, key dates and positions of a matter carry across conversations, according to the vendor.
- Drafting and Word: AI revisions appear as tracked changes, and there is a Word add-in in a side panel.
- Connections: import from Google Drive and OneDrive, and use CaseRead from Claude and ChatGPT, which the vendor describes as read-only.
Where its law comes from
The sources page lists CaseRead's own public law database for statutes and regulations in federal law, all 50 states, DC and Puerto Rico, the Free Law Project for case law, and eCFR, the Federal Register and GovInfo for federal publications. The Federal Rules of Civil Procedure and of Evidence are kept on hand, and state court rules vary by state.
The terms add that AI from Anthropic and OpenAI helps with research, drafting and analysis, and that results from outside data sources are provided as is and may be incomplete or out of date.
Who can sign up
The terms say a user is a licensed attorney in good standing in at least one US jurisdiction, a paralegal or staff member supervised by one, or a self-represented person who accepts responsibility for verifying AI output before relying on it. Every firm gets its own database schema, the vendor says, and nothing uploaded is used to train a model. These are the vendor's statements, labeled as such.
Security claims on the home page include encryption in transit and at rest, two-factor sign-in that a firm administrator can require, hosting in the United States and a firm activity log. The vendor describes the setup as aligned with SOC 2, which is a different thing from a completed audit report, so ask for documents if your firm needs them.
What it is not, and where to go next
- It is not legal advice, and using it creates no attorney-client relationship with the vendor, according to its terms.
- It is not a replacement for checking that a case is still good law in a citator.
- It is not the only route. The solo attorney guide lists others, and the review covers the limits.
- To check a draft by hand instead, follow the citation checking steps.
Sources used for this page
These records support the facts and comparisons above. Merchant-controlled records are labelled so you can separate product claims from independent evidence.
- CaseRead home page — Merchant documentation · caseread.ai · Merchant-controlled · checked 2026-11-04
- CaseRead terms of service — Merchant documentation · caseread.ai · Merchant-controlled · checked 2026-11-04
- CaseRead Hallucination Shield page — Merchant documentation · caseread.ai · Merchant-controlled · checked 2026-11-04
- CaseRead sources page — Merchant documentation · caseread.ai · Merchant-controlled · checked 2026-11-04
- CaseRead comparison with ChatGPT — Merchant documentation · caseread.ai · Merchant-controlled · checked 2026-11-04