EARLY WARNING FOR REGULATED PROFESSIONALS
See the court cases, rulings and rules on AI in professional work, find what is relevant to your profession, and act before the risk reaches your firm. Every figure is calculated from the source data, and every case and development links to its source.
1 · THE PROBLEM
COURT CASES TRACKED
2,100
Court cases involving AI-fabricated content across 43 jurisdictions since 2023
231% growth in the last 12 months
October 2025 to September 2026, against the previous 12 months. See method and sources
1 of 2,100 cases verified by Ecsper against court records
LATEST LANDMARK State v Horcasitas, Arizona, September 2026: Sentence vacated
2 · THE RESPONSE
NEXT AND LATEST RULES AND RULINGS
| Date | Rule or ruling | Region |
|---|---|---|
| Nov 2026 Scheduled | RICS expert-witness standard takes effect | UK |
| Sep 2026 | Ireland High Court AI practice direction HC 142 comes into operation | EU |
| Aug 2026 | Kumar: removal from the Register of Foreign Lawyers | UK |
| Aug 2026 | AI Act transparency and further enforcement commence | EU |
27 since 2023; 6 more scheduled. See the timeline
By profession
Relevance, not a compliance conclusion
3 · THE GAP
91%
of professionals use AI at work
7%
keep audit logs of their AI use
34%
use AI their firm has not authorised
32%
of clients are reconsidering firms without AI governance.
EXPERT WITNESSES
17%
of firms use AI to support expert witness work
45%
plan to start within 12 months
12%
have a written policy for AI in expert work
3%
check whether AI outputs would survive a court challenge
How the numbers are changing.
4.3
new cases a day, on average over the same 12 months
'Not published' indicates cases where the court did not publish a monetary penalty. This does not mean no penalty was imposed.
These are only the cases a court caught and wrote up. The true number is higher.
Explore the evidence
Filterable, sortable, and searchable.
Total cases Loading
Cases with an identified AI tool Loading
Most common tool Loading
Tap a card to expand details
The rulings that changed practice
8 cases that have shaped how courts, regulators, and the profession respond to AI in professional work.
Landmark cases include rulings beyond fabricated citations, such as orders on the disclosure of AI prompts. The Case Explorer covers AI-fabricated court filings only.
September 2026
At a manslaughter sentencing, the judge relied on an AI-generated video of the victim speaking. The Arizona Court of Appeals held the video was not reliable, found fundamental error and vacated the sentence for resentencing.
September 2026
In the Riverside County ballot-seizure case, a law firm misquoted a Supreme Court precedent. The California Supreme Court disapproved the improper use of AI tools that apparently contributed, reminded counsel of their duties of competence and candour, and referred the matter to the State Bar.
August 2026
3M's expert asked ChatGPT to "show how 3M is 0% at fault". Plaintiffs obtained hundreds of pages of his prompts in discovery and called him as their own witness. A Harris County jury found 3M 30% liable and awarded $61.5M. 3M has said it will appeal.
404 Media, Forbes, CBS News
June 2026
A prosecution witness used AI to generate likely cross-examination questions and model answers before giving evidence. The Court of Appeal treated this as witness coaching but reversed the stay of proceedings, and said witnesses should be strongly discouraged from using AI to prepare their evidence.
May 2026
The expert used GPT-4o to filter Shell's documents. A magistrate judge ruled the prompts are part of her methodology and discoverable under Rule 26, reported as the first federal order of its kind. The district court stayed the order in June pending CLF's objection.
Mayer Brown, Dechert, Arnold & Porter
June 2026
Both sides filed AI-fabricated citations. The trial was cancelled, all four lawyers were removed from the case, the two out-of-state lawyers lost their admission and were barred from the district for two years, and fines totalled $8,000.
Sanctions order (N.D. Miss.), Bloomberg Law
December 2025 to March 2026
Oregon winery dispute. Three briefs cited 15 non-existent cases and eight fabricated quotations. The court dismissed the claims with prejudice and awarded sanctions and fees of $110,204.38, reported as the largest US penalty for AI-fabricated filings.
ABA Journal, D. Oregon orders
June 2023
Lawyer asked ChatGPT for cases. Six did not exist. $5,000 sanction. Three years later, the same failure has its own public database and the database needs daily updates.
Court filings
AI governance regulations, guidance, and court rulings across jurisdictions. Tracking the shift from advised to enforced.
No events match these filters.
Select a milestone for detail. Dashed outlines indicate scheduled dates. Sources: public regulatory publications, court records, and professional body guidance.
Provenance and interpretation
All figures in Ecsper AI Risk Intelligence are drawn from independently published sources. This page documents the provenance, scope, and update schedule for each.
AI Hallucination Cases Database
Includes data from the AI Hallucination Cases Database by Damien Charlotin (https://www.damiencharlotin.com/hallucinations/), licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). Ecsper adds its own verification against court records, references and analysis. Verified by Ecsper: we opened the court record ourselves and recorded its reference. Source-reported: recorded in the source database and not yet checked by us. The figures on this page are drawn from the AI Hallucination Cases database maintained by Damien Charlotin, a researcher at Sciences Po, Paris. The database is publicly available, licensed under CC BY 4.0 with attribution, and its inclusion bar is strict: only decisions where a court or tribunal explicitly found or clearly implied reliance on hallucinated material. Figures are calculated live from the case records. The source database is refreshed every 14 days. This tracker counts documented outcomes. Cases where fabricated content was filed but not caught are, by definition, invisible. Whatever the true number is, this is the floor. The full dataset is available as a CSV download from the source database. Ecsper downloads, trims, and analyses the dataset monthly. The case explorer shows the trimmed dataset; for the complete record including hallucination item details and source documents, visit the original database.
Thomson Reuters Institute, 2026
Contextual statistics on AI adoption and governance gaps are drawn from the Thomson Reuters Institute Future of Professionals Report 2026, based on a survey of 1,816 professionals across 62 countries. These figures appear in the Overview dashboard. They are cited individually where they appear and are not drawn from the Charlotin database.
Secretariat and ACEDS, 2026
Expert witness AI adoption and controls data are drawn from the Secretariat and ACEDS 2026 Artificial Intelligence Report, a survey of legal professionals conducted in March 2026 across law firms, corporations, government agencies, service providers, and independent consultancies. This is the first year the survey included a dedicated section on AI use with expert witnesses.
Stanford RegLab, 2024
Error rates for commercial legal AI tools are drawn from a peer-reviewed Stanford RegLab study, which found hallucination rates of 17% to 34% even in paid platforms including Westlaw AI-Assisted Research and Lexis+ AI. This finding is cited in the growth trajectory section.
Blue identifies adoption indicators; red identifies risk and control indicators. Percentages from separate surveys are not additive. Court cases, lawyers, and sanctions are different counting units.
Includes data from the AI Hallucination Cases Database by Damien Charlotin (https://www.damiencharlotin.com/hallucinations/), licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). Ecsper adds its own verification against court records, references and analysis.
Verified by Ecsper: we opened the court record ourselves and recorded its reference. Source-reported: recorded in the source database and not yet checked by us.
Thomson Reuters Institute. Future of Professionals Report 2026.
Dahl, M. et al. 'Hallucination-Free? Assessing the Reliability of Leading AI Legal Research Tools.' Stanford RegLab, 2024.
Secretariat and ACEDS. 2026 Artificial Intelligence Report. July 2026.