A map of the field
| Cluster | What is disputed | Example source |
|---|---|---|
| Fabricated legal authorities | Verification duties and sanctions | Mata v. Avianca |
| Training data and copyright | Protected editorial content and fair use | Thomson Reuters v. ROSS |
| AI-generated evidence | Authentication, reliability and foundation | U.S. Courts overview |
| Product representations | Who bears loss when automated statements are wrong | Jurisdiction-specific consumer and contract law |
How we select cases
We link to an order, official docket or government publication wherever one is available. A complaint is labelled as an allegation; an interim ruling is not described as a final result; an appeal is not treated as resolved. We do not inflate the tracker by repeating syndicated stories about the same proceeding.
Why the cases matter to judges and lawyers
AI litigation is also operational guidance. Citation sanctions show the need for source verification. Copyright disputes expose provenance and licensing questions. Evidence debates show why a generated image, transcript or enhancement needs a foundation. Product cases remind organisations that a chatbot can create representations even when no human intended the exact words.
Current verified entries
Mata v. Avianca. The Southern District of New York imposed sanctions after submissions included non-existent cases and false quotations. The order focuses on conduct and verification, not a blanket ban on AI.
Thomson Reuters v. ROSS Intelligence. The District of Delaware litigation concerns Westlaw headnotes and material used in developing a competing legal research tool. It is a copyright case, not a judicial-drafting case.
Updates appear in Court AI News and Research after source review.