Concepts (6)
Expert witness
Concept
A witness permitted to state an opinion because of specialized knowledge, whose methodology is open to examination.
An expert's methodology has always been open to examination. That is the condition on which opinion testimony is received at all.
The Expert Record Standard — Hartzer Press
Artificial intelligence
Concept
Systems that generate content, analysis, summaries, or conclusions in response to a request.
In this standard, system means any artificial intelligence system that generates content, analysis, summaries, or conclusions in response to a request, whether accessed directly or operating inside other software.
The Expert Record Standard — Hartzer Press
Expert report
Concept
The written statement of a retained expert's opinions and their bases.
Nothing enters the report on the authority of a system. Every citation, figure, date, quotation and specification is confirmed against a source that exists independently of the system.
The Expert Record Standard — Hartzer Press
Maintained reference
Taxonomy
Sections of this site kept current with a checked-on date.
The case record, the working papers and errata are maintained here because they change; the book carries the method, which does not.
The Expert Record — Hartzer Press
Hallucination (artificial intelligence)
Concept
Output from an AI system presented as real that does not exist, such as a citation to a case that was never decided.
This part of the record collects matters in which a filing put something before a tribunal that did not exist: a case, a quotation, a statutory passage or a fact from the trial record. In each matter recorded here, the fabricated material came from, or was attributed to, a generative AI system (software that produces text in response to a prompt).
Fabricated citations and facts in filings — Hartzer Press
Deepfake
Concept
Synthetic audio, images or video made to depict a real person saying or doing something they did not.
This part of the record follows matters in which the evidence itself was produced or changed by an AI system, or in which a party said it had been. It is distinct from matters about lawyers or experts using AI to write briefs and reports.
AI-generated and altered evidence — Hartzer Press
The book and the Standard (7)
The Expert Record Standard
Concept
Eight clauses governing the use of artificial intelligence in expert witness engagements, licensed for verbatim adoption.
An engagement satisfies this standard when a competent examiner in the same field, working only from the primary sources, with no access to any artificial intelligence system, could reproduce the expert's result.
The Expert Record Standard — Hartzer Press
theexpertrecord.com
WebSite
The Expert Record by Bill Hartzer: the book and the eight-clause Standard for expert witnesses using AI, with the companion case record and working papers.
Artificial intelligence did not create a new way for an expert to be wrong. It created a new record of how the expert got there. That record is complete, timestamped, verbatim, and held by someone else, and it is increasingly producible in discovery. This book is the working method for making sure yours would survive being read aloud.
The Expert Record — The Standard for Expert Witnesses Using AI — Hartzer Press
The Expert Record
Book
The Expert Record: The Standard for Expert Witnesses Using AI, by Bill Hartzer. Thirty-two chapters, the eight clauses, and nine appendices.
The Standard for Expert Witnesses Using AI. Thirty-two chapters, the eight clauses, and nine appendices.
The Expert Record: the book — Hartzer Press
Adopting the Standard
CreativeWork
License terms and attribution for the Expert Record Standard, and where it belongs in an engagement letter.
What the license permits, how to attribute it, and where it goes in an engagement.
Adopting the Standard | The Expert Record — Hartzer Press
Verification Log Fields and the Standard
CreativeWork
The fields a source-verification log should record, each mapped to the clause of the Expert Record Standard v1.0 that requires it.
The fields a source-verification log should record, and the clause of the Standard each one serves.
Verification Log Fields Mapped to the Standard | TER — Hartzer Press
Before You Use AI on an Engagement
CreativeWork
A checklist for expert witnesses before, during and after using AI on an engagement, with each item tied to a clause of the Expert Record Standard.
17 decisions, in the order an engagement meets them, each tied to the clause of the Standard it serves. Print it, adopt it, or keep it beside the log.
Before You Use AI on an Engagement: A Checklist | TER — Hartzer Press
How this site is researched
CreativeWork
How theexpertrecord.com is researched: sources, how each fact is traced, the part AI systems play, review before publication, and dating.
Where the material comes from, how each fact is traced to a document, the part AI systems play in the work, and how pages are dated and corrected.
How This Site Is Researched | The Expert Record — Hartzer Press
The Standard, clause by clause (8)
Clause 1: Sequence
Concept
Why the Expert Record Standard puts sequence first, what the clause forbids and permits, how to prove the order of the work, and what the record shows.
Form the conclusion first. Never ask a system what the evidence supports.
Clause 1, Sequence: Conclusion Before the System | TER — Hartzer Press
Clause 2: Non-delegation
Concept
/standard/clause-2-non-delegation/
What non-delegation protects in expert work, what it leaves permitted, how to show authorship, and how tribunals have treated opinions an AI tool produced.
The opinion and the decision to give it are yours alone.
Clause 2, Non-delegation: The Opinion Stays Yours | TER — Hartzer Press
Clause 3: Independent authority
Concept
/standard/clause-3-independent-authority/
Independent authority explained: why nothing enters an expert report on an AI tool's word, how a verification pass works, and what the citation record shows.
Nothing enters the report on a system's say-so.
Clause 3, Independent Authority: What It Requires | TER — Hartzer Press
Clause 4: Checkability before adoption
Concept
/standard/clause-4-checkability-before-adoption/
Checkability before adoption: the one-sentence test an expert applies before accepting AI output, the misreadings to avoid, and what tribunals have done.
If you cannot name what would check it, do not use it.
Clause 4, Checkability: Name the Check First | The Expert Record — Hartzer Press
Clause 5: Reproducible analysis
Concept
/standard/clause-5-reproducible-analysis/
Reproducible analysis under the Expert Record Standard: why counts and selections need a re-runnable method, what to keep, and what the record shows.
Numbers come from a method that repeats, and you keep the method.
Clause 5, Reproducible Analysis: Keep the Method | TER — Hartzer Press
Clause 6: Custody and confidentiality
Concept
/standard/clause-6-custody-and-confidentiality/
Custody and confidentiality for experts using AI: why an upload is a disclosure, what to settle before material arrives, and orders that put a price on it.
Know what leaves your environment, and who keeps it.
Clause 6, Custody and Confidentiality Explained | TER — Hartzer Press
Clause 7: Contemporaneous record
Concept
/standard/clause-7-contemporaneous-record/
The contemporaneous record explained: why the AI-use log makes every other clause provable, what an entry contains, and what tribunals asked without one.
Kept as you work, not assembled afterward.
Clause 7, Contemporaneous Record: Log as You Work | TER — Hartzer Press
Clause 8: Disclosure
Concept
/standard/clause-8-disclosure/
Disclosure under the Expert Record Standard: why the commitment comes at retention, how much a report says, and what candor and denial have cost.
Said at retention, honored on request.
Clause 8, Disclosure: Say It at Retention | The Expert Record — Hartzer Press
Guides and glossary (7)
Guides to AI in the Courtroom
Taxonomy
Guides to AI in the Courtroom: proposed Rule 707, discovery of AI prompts, disclosure rules, deepfake claims and expert reports citing nonexistent sources.
Cross-cutting questions about AI in litigation, each answered from the matters in the record and linked to every one of them.
Guides to AI in the Courtroom | The Expert Record — Hartzer Press
AI Disclosure and Certification Rules, Compared
CreativeWork
/ai-in-court/guides/ai-disclosure-and-certification-rules/
Court rules, standing orders, statutes and guidance that require disclosure of AI use or certification of filings, compared by status, scope and consequence.
Fifteen instruments in the record ask filers either to say that AI was used or to certify that its output was checked. They differ on who carries the duty, what it covers and whether experts are reached at all.
AI Disclosure and Certification Rules Compared | TER — Hartzer Press
Deepfake Claims About Evidence: How Tribunals Have Ruled
CreativeWork
/ai-in-court/guides/deepfake-claims-about-evidence/
How tribunals have handled evidence alleged to be AI-generated or altered, and claims that genuine recordings are deepfakes, with federal rulemaking status.
Fabricated exhibits and the claim that real evidence is fabricated have both reached tribunals. The responses differ, and the record shows what each tribunal asked for.
Deepfake Claims About Evidence and Authentication | TER — Hartzer Press
Discovery of AI Prompts and Chat Logs: What the Record Shows
CreativeWork
/ai-in-court/guides/discoverability-of-ai-prompts-and-chat-logs/
How tribunals have treated requests for experts' and parties' AI prompts, chat logs and outputs, with privilege, work product and each order's posture.
A small, growing set of orders addresses whether an expert's or a party's exchanges with a generative system must be produced. The outcomes differ, and nearly all are trial-level rulings.
Discovery of AI Prompts and Chat Logs in Litigation | TER — Hartzer Press
Expert Reports That Cite Sources That Do Not Exist
CreativeWork
/ai-in-court/guides/expert-reports-citing-sources-that-do-not-exist/
What tribunals did when an expert report or declaration relied on sources that do not exist: struck, excluded, fee zeroed, discounted or left to cross.
Eighteen matters in the record show what tribunals did when expert material cited sources that could not be found or output nobody could verify. The outcomes differ, and the differences are instructive.
Expert Reports Citing Sources That Do Not Exist | TER — Hartzer Press
Proposed Rule 707: AI Evidence and the Daubert Standard
CreativeWork
/ai-in-court/guides/proposed-rule-707-ai-evidence-daubert/
Proposed FRE 707 would hold AI output offered without an expert to Rule 702 and Daubert. The text, the comments, the revised draft, and why it is not law yet.
Many articles say AI evidence must now meet Daubert under a new Federal Rule of Evidence 707. The proposal is real, but it was withdrawn in the form published and a revised draft is still under study. This guide sets out the text, the record behind it and where it stands.
Proposed Rule 707: AI Evidence and the Daubert Test | TER — Hartzer Press
Glossary
DefinedTermSet
Plain definitions of the AI, evidence and procedure terms used on The Expert Record, from hallucination and deepfake to work product and the governing test.
Thirty-five terms used across this site, each defined plainly, with the legal system a rule belongs to and a link to where the term matters in the record.
Glossary: AI, Evidence and Expert Witness Terms | TER — Hartzer Press
Case record and corrections (2)
The case record
Dataset
Every matter described in The Expert Record, identified in full, with disposition, current posture, and the date each entry was last checked.
Every matter the book describes, identified in full, with the disposition, the current posture, and the date each entry was last checked.
The case record | The Expert Record — Hartzer Press
Errata and updates
CreativeWork
Corrections to The Expert Record, and notice of any material in it that has been superseded.
Corrections to the book, and a note whenever something in it has been superseded.
Errata and updates | The Expert Record — Hartzer Press
Working papers (5)
The working papers
Taxonomy
The AI-use log template, engagement letter provisions, verification checklist, deposition question sets, analysis specifications, and preservation protocol.
The six working papers behind the book: three published here in full, all six summarized.
The working papers | The Expert Record — Hartzer Press
AI-use log template
CreativeWork
The log template from Appendix A of the book, in full: the header block, the per-use rows, and the four rules that make a log worth having.
Reproduced in full from the book, so that a template can become your template.
AI-use log template | The Expert Record — Hartzer Press
Preservation and production protocol
CreativeWork
/working-papers/preservation-protocol/
The preservation protocol from Appendix F: the day-one folder, the export cadence, what to do when a request arrives, and the cross-matter problem.
Reproduced in full from the book, so that a template can become your template.
Preservation and production protocol | The Expert Record — Hartzer Press
Analysis scripts
CreativeWork
Runnable implementations of the seven reproducible-analysis specifications from Appendix E, maintained against current language versions.
The book carries the specification. The code lives here, because code ages and a listing printed in a book is a listing nobody can run in three years.
Analysis scripts | The Expert Record — Hartzer Press
The verification checklist
CreativeWork
/working-papers/verification-checklist/
The verification checklist from Appendix C: the six categories of checkable assertion, the four dispositions, and what to do when a source cannot be located.
Reproduced in full from the book, so that a template can become your template.
The verification checklist | The Expert Record — Hartzer Press
AI in the Courtroom: index and topic hubs (7)
AI in the Courtroom
Taxonomy
A maintained, dated record of real matters in which an AI system was used in litigation, grouped into six topics.
AI in the Courtroom records matters in which an artificial intelligence system was used in litigation and a court, a rulemaker or a regulator then did something about it. It covers lawyers' filings, expert reports, the evidence itself, judges' own orders, what was shown or said in a courtroom, and the rules written in response.
AI in the Courtroom — Hartzer Press
Expert witnesses and AI
Taxonomy
/ai-in-court/expert-witnesses/
Expert declarations and reports prepared with an AI system, and how tribunals treated the evidence. 21 matters, each dated and sourced.
Expert declarations and reports prepared with an AI system, and how tribunals treated the evidence.
Expert witnesses and AI: AI in the Courtroom | The Expert Record — Hartzer Press
Fabricated citations and facts in filings
Taxonomy
/ai-in-court/fabricated-authority/
Filings that put a nonexistent case, quotation or fact before a court, and what the court did about it. 32 matters, each dated and sourced.
Filings that put a nonexistent case, quotation or fact before a court, and what the court did about it.
Fabricated citations and facts in filings | The Expert Record — Hartzer Press
AI-generated and altered evidence
Taxonomy
Deepfakes offered as evidence, AI-enhanced media, and claims that genuine recordings were fabricated. 20 matters, each dated and sourced.
Deepfakes offered as evidence, AI-enhanced media, and claims that genuine recordings were fabricated.
AI-generated and altered evidence: AI in the Courtroom | TER — Hartzer Press
Judges, courts and AI
Taxonomy
/ai-in-court/judges-and-courts/
Orders containing AI errors, judges using AI openly, and the oversight that followed. 20 matters, each dated and sourced.
Orders containing AI errors, judges using AI openly, and the oversight that followed.
Judges, courts and AI: AI in the Courtroom | The Expert Record — Hartzer Press
AI inside the courtroom
Taxonomy
/ai-in-court/in-the-courtroom/
Avatars, generated statements, AI in place of counsel, and AI tools used in proceedings. 7 matters, each dated and sourced.
Avatars, generated statements, AI in place of counsel, and AI tools used in proceedings.
AI inside the courtroom: AI in the Courtroom | The Expert Record — Hartzer Press
Rules, orders and guidance
Taxonomy
/ai-in-court/rules-and-guidance/
Rules of evidence, court rules, standing orders, ethics opinions and expert-facing guidance on AI. 13 matters, each dated and sourced.
Rules of evidence, court rules, standing orders, ethics opinions and expert-facing guidance on AI.
Rules, orders and guidance: AI in the Courtroom | TER — Hartzer Press
AI in the Courtroom: Expert witnesses and AI (21)
In re Celsius Network LLC (Faraj report)
Event
/ai-in-court/in-re-celsius-network-faraj-report/
A bankruptcy court excluded a 172-page CEL token valuation report written by AI in 72 hours but admitted the expert's live testimony.
An objector's valuation expert in the Celsius bankruptcy directed an AI system to write a 172-page report in 72 hours. The bankruptcy court excluded the report as unreliable and admitted his live testimony.
In re Celsius: AI-Written Expert Report Excluded | TER — Hartzer Press
Matter of Weber
Event
/ai-in-court/matter-of-weber-copilot-damages/
A New York Surrogate's Court gave no weight to damages figures an expert cross-checked with Microsoft Copilot and set a disclosure rule for AI evidence.
A damages expert in a New York trust accounting used Microsoft Copilot to cross-check his calculations and could not recall his prompts. The Surrogate's Court rejected his analysis and announced a disclosure and Frye hearing requirement for AI-generated evidence.
Matter of Weber: Copilot Damages Calculations | The Expert Record — Hartzer Press
Plonit v. Clal Insurance (DigitalOwl summary)
Event
/ai-in-court/plonit-v-clal-digitalowl-expert/
An Israeli court ordered a court-appointed medical expert to disregard an AI-processed claim summary and made the insurer resend the records unmarked.
An insurer ran a claimant's medical records through an AI analysis platform and sent the resulting summary to the court-appointed expert. The court ordered the expert to disregard it and the records resent without markings.
Plonit v. Clal: AI Summary Kept From Expert | The Expert Record — Hartzer Press
Kohls v. Ellison
Event
/ai-in-court/kohls-v-ellison-hancock-declaration/
A Minnesota federal court excluded an expert declaration drafted with GPT-4o after 2 fabricated articles and 1 misattributed citation were found.
An expert defending Minnesota's election deepfake law used GPT-4o to help draft his declaration, which cited 2 articles that do not exist. The court excluded the declaration and denied leave to file a corrected one.
Kohls v. Ellison: Expert Declaration Excluded | The Expert Record — Hartzer Press
NSW Supreme Court Practice Note SC Gen 23
Event
/ai-in-court/nsw-practice-note-expert-reports-gen-ai/
Since February 2025 the NSW Supreme Court has barred generative AI in expert reports without leave, and requires disclosure and records if allowed.
The Supreme Court of New South Wales issued a practice note barring the use of generative AI to draft or prepare expert reports without prior leave. Where leave is given, the expert must disclose the use and annex a record of it.
NSW Practice Note SC Gen 23: AI in Expert Reports | TER — Hartzer Press
Ferlito v. Harbor Freight Tools USA
Event
/ai-in-court/ferlito-v-harbor-freight-chatgpt/
An E.D.N.Y. judge denied a motion to exclude a product-design expert who used ChatGPT only after finishing his report, to confirm his conclusion.
A plaintiff's product-design expert asked ChatGPT a question after completing his report, and the answer matched his opinion. The defendant moved to exclude him on that basis. The court denied the motion.
Ferlito v. Harbor Freight: ChatGPT Check Allowed | TER — Hartzer Press
Concord Music Group v. Anthropic (Chen declaration)
Event
/ai-in-court/concord-music-v-anthropic-chen-declaration/
Paragraph 9 of Olivia Chen's expert declaration was struck after Claude.ai, used by counsel to format a citation, produced a wrong title and wrong authors.
Counsel used Claude.ai to format citations in a sampling expert's declaration, and it produced a wrong title and wrong authors for a real article. The magistrate judge struck paragraph 9 and found the error undermined the declaration's credibility.
Concord v. Anthropic: The Chen Declaration | The Expert Record — Hartzer Press
Butler and National Disability Insurance Agency
Event
/ai-in-court/butler-ndia-ai-assisted-expert-reports/
A tribunal gave reduced weight to AI-assisted expert reports, holding a practitioner must remain responsible and accountable for a report's content.
Expert reports prepared with AI assistance were given reduced weight by an Australian tribunal, which held that a practitioner may use a system to reduce personal observations to writing but must remain accountable for the content.
Butler: AI-Assisted Expert Reports Given Less Weight | TER — Hartzer Press
In re Valsartan MDL (Dr. William Sawyer)
Event
/ai-in-court/valsartan-mdl-sawyer-ai-citations/
A special master in the Valsartan MDL declined to exclude a toxicologist whose report cited non-existent sources produced by an AI tool.
A plaintiff's toxicology expert in the Valsartan MDL cited non-existent sources after using an AI tool without verifying them. The special master denied the motion to exclude and left the issue to costs and cross-examination.
Valsartan MDL: Expert Kept Despite AI Citations | TER — Hartzer Press
Biton v. The National Insurance Institute
Event
/ai-in-court/biton-national-insurance-chatgpt-expert/
A labor court let a claimant put ChatGPT-sourced research to a court-appointed expert, but required the articles themselves, not the system's summaries.
A claimant who could not afford a physician used ChatGPT to research his injury and sought to question the court-appointed expert about it. The court allowed it, but required the underlying articles rather than the system's summaries.
Biton: ChatGPT Sources Put to a Court Expert | The Expert Record — Hartzer Press
Landgericht Darmstadt expert fee decision
Event
/ai-in-court/lg-darmstadt-expert-fee-zero/
A German regional court set a court-appointed expert's fee at zero euros after finding his report was AI-generated and the claimant was never examined.
A court-appointed medical expert submitted a report the court found AI-generated. The court set his fee at zero euros rather than the amount claimed, on the ground that a report is compensable only insofar as it is usable.
Expert Fee Set to Zero for an AI Report | The Expert Record — Hartzer Press
Cour d'appel de Rouen, RG 25/01677 (Copilot analysis)
Event
/ai-in-court/rouen-court-of-appeal-copilot-analysis/
In a drainage dispute, a French appeals court relied on professional evidence and expressly not on a party's Copilot analysis of a contractor's letter.
Homeowners suing neighbors over rainwater infiltration offered a Microsoft Copilot analysis of the contractor's written statement as support for causation. The Rouen Court of Appeal found for them, but rested causation on the agreement of professional evidence and expressly excluded the AI analysis from that evidence.
Rouen Court of Appeal: A Copilot Analysis Not Relied On | TER — Hartzer Press
Social Media MDL (Osborne AI citation tool)
Event
/ai-in-court/social-media-mdl-osborne-citations/
In the social media adolescent addiction MDL the court declined to exclude a school-district expert over citations attributed to an AI citation tool.
Defendants argued that a school-district expert relied on articles that do not exist and that other citations were AI-generated miscites. The court declined to exclude the opinions and left the issue to cross-examination.
Social Media MDL: Miscites Left to Cross-Examination | TER — Hartzer Press
Riddle Aggregates (Tax Court rebuttal expert)
Event
/ai-in-court/riddle-aggregates-tax-court-appraisal/
The Commissioner withdrew two rebuttal appraisal reports mid-trial in the U.S. Tax Court after voir dire exposed citations to authority that does not exist.
The Commissioner's rebuttal appraisal expert acknowledged using ChatGPT and Gemini. Voir dire established that his reports cited nonexistent authority. After the judge stated grave concerns and barred further direct examination, the Commissioner withdrew both reports.
Tax Court: A Government Expert's Reports Withdrawn | TER — Hartzer Press
Keratoconus Entitlement Review, 2026 CanLII 107034
Event
/ai-in-court/vrab-keratoconus-ai-medical-opinion/
A Canadian veterans' review panel found a clinician's opinion cited a nonexistent agency guideline and two untraceable articles, and denied entitlement.
A treating nurse practitioner's opinion letter supporting a veteran's claim cited a Veterans Affairs Canada guideline that does not exist and two journal articles the panel could not find. The Veterans Review and Appeal Board found the clinician had used AI, held the opinion not credible and denied entitlement.
Keratoconus Review: AI-Built Medical Opinion Rejected | TER — Hartzer Press
Conservation Law Foundation v. Shell (expert AI prompts)
Event
/ai-in-court/clf-v-shell-expert-ai-prompts/
A Connecticut magistrate judge ordered production of the AI prompts an expert's team used to cull a document production. The order is stayed on objection.
An expert's team used a generative AI system to cull the defendants' document production. A magistrate judge ordered the prompts disclosed as part of the expert's methodology. The order is stayed while an objection is decided.
CLF v. Shell: An Expert's AI Prompts Ordered Produced | TER — Hartzer Press
Bard Port Catheter MDL (El-Ghannam ChatGPT challenge)
Event
/ai-in-court/bard-port-catheter-mdl-el-ghannam/
An MDL court declined to exclude a materials-science expert who acknowledged using ChatGPT, where no hallucinated citation was identified in his report.
A plaintiffs' materials-science expert acknowledged at deposition that he used ChatGPT among other tools. The court declined to reach the AI argument, noting it was raised for the first time on reply and that no fabricated citation had been identified in the report at issue.
Bard MDL: Suspicion of AI Use Is Not Enough | The Expert Record — Hartzer Press
LeDoux v. Outliers, Inc.
Event
/ai-in-court/ledoux-v-outliers-ai-citation-table/
A Tacoma federal judge fined counsel $3,000 over AI-generated citations, including a table given to 2 experts, and later excluded a third expert's report.
Plaintiff's counsel used an AI system to generate a citation table and gave it to two experts, who attached it to their reports. The court personally sanctioned counsel $3,000 and later excluded a third expert whose report contained fabricated citations.
LeDoux v. Outliers: AI Citation Table in Expert Reports | TER — Hartzer Press
Watson Grinding explosion trial (3M expert)
Event
/ai-in-court/watson-grinding-3m-expert-chatgpt/
A 3M expert in a Houston explosion trial produced about 350 pages of ChatGPT logs and was questioned at trial about a report largely drafted by ChatGPT.
An engineering expert retained by 3M in Houston explosion litigation used ChatGPT to draft most of his report and produced about 350 pages of chat logs. He was examined on them at trial. The jury found 3M 30% responsible.
Watson Grinding Trial: 3M Expert's ChatGPT Logs | TER — Hartzer Press
Villanueva v. Las Vegas Metropolitan Police Department
Event
/ai-in-court/villanueva-v-lvmpd-expert-chatgpt-log/
A federal judge ordered an expert's ChatGPT case summaries produced as facts or data considered, and held his prompts and draft-report output protected.
The plaintiff's expert in an excessive-force case used ChatGPT to research and draft his report and disclosed it only at deposition. The district judge ordered the ChatGPT case-law summaries produced as facts or data considered, and held his prompts, a deposition outline and draft-report output protected.
Villanueva: What an Expert Must Produce From ChatGPT | TER — Hartzer Press
Avergil v. Yaakov
Event
/ai-in-court/avergil-v-yaakov-expert-report-struck/
An Israeli court struck a banking expert's report that cited a nonexistent precedent, an untraceable book and invented text of the Payment Services Law.
A banking expert's report filed for a lawyer-defendant in an Israeli negligence suit relied on a nonexistent Supreme Court precedent, a misattributed decision, an untraceable book and invented statutory text. The magistrates' court struck the report before trial, refused a related amendment and awarded costs.
Avergil v. Yaakov: Expert Report Struck Over AI Errors | TER — Hartzer Press
AI in the Courtroom: Fabricated citations and facts in filings (32)
Mata v. Avianca, Inc.
Event
Two lawyers filed a brief citing airline decisions ChatGPT invented, then stood by them under questioning. What the court did, and what it cost them.
Plaintiff's lawyers in a personal injury suit against Avianca filed a brief citing cases produced by ChatGPT that did not exist. Judge P. Kevin Castel imposed a $5,000 penalty and ordered letters to the client and to the judges named as authors of the fake opinions.
Mata v. Avianca: The ChatGPT Fake Cases | The Expert Record — Hartzer Press
Ayinde v Haringey and Al-Haroun v Qatar National Bank
Event
/ai-in-court/ayinde-al-haroun-divisional-court/
The Divisional Court of England and Wales referred a barrister and solicitors to regulators in June 2025 after filings cited cases that do not exist.
In 2 cases heard together, lawyers in England put fictitious case law before the courts. The Divisional Court declined contempt proceedings, referred a barrister and 2 solicitors to regulators and sent its judgment to the profession's governing bodies.
Ayinde and Al-Haroun: High Court on Fake Citations | TER — Hartzer Press
Johnson v. Dunn (Butler Snow)
Event
/ai-in-court/johnson-v-dunn-butler-snow/
N.D. Ala. publicly reprimanded and disqualified 3 Butler Snow lawyers and referred them to the bar after ChatGPT supplied 5 false citations in 2 motions.
Lawyers at Butler Snow defending Alabama prison officials filed 2 motions containing 5 citations a partner obtained from ChatGPT that did not exist or did not say what was claimed. The court reprimanded 3 lawyers, disqualified them and referred them to the bar.
Johnson v. Dunn: Butler Snow Lawyers Disqualified | TER — Hartzer Press
Victorian murder trial submissions (Nathwani KC)
Event
/ai-in-court/nathwani-victoria-murder-case/
Defense submissions in a Supreme Court of Victoria murder case contained AI-generated fake quotes and citations. Senior counsel apologized in August 2025.
Defense counsel in a Victorian murder case filed submissions containing fabricated quotations and nonexistent case citations generated by AI. Rishi Nathwani KC apologized. Justice James Elliott called the events unsatisfactory and ruled a day later than planned.
AI-Generated Submissions in a Victorian Murder Trial | TER — Hartzer Press
Noland v. Land of the Free, L.P.
Event
/ai-in-court/noland-v-land-of-the-free/
California Court of Appeal fined counsel $10,000 and referred him to the State Bar after 21 of 23 case quotations in his opening brief proved fabricated.
An appellant's lawyer in a California employment appeal filed briefs whose case quotations were largely fabricated by AI tools. The Court of Appeal affirmed the judgment, sanctioned him $10,000 and referred him to the State Bar.
Noland v. Land of the Free: Appellate AI Sanction | TER — Hartzer Press
Kjoller v. Superior Court (Nevada County DA)
Event
/ai-in-court/kjoller-nevada-county-prosecutor/
California Supreme Court directed an order to show cause on sanctions after Nevada County prosecutors filed briefs citing authority that does not exist.
Prosecutors in Nevada County, California filed briefs in felony cases that cited authority that does not exist. The California Supreme Court directed the Court of Appeal to issue an order to show cause on sanctions, and a judge was appointed in 2026 to investigate.
Nevada County Prosecutors and AI Citations: Kjoller | TER — Hartzer Press
Amarsingh v. Frontier Airlines (Tenth Circuit)
Event
/ai-in-court/amarsingh-v-frontier-tenth-circuit/
The Tenth Circuit ordered a pro se appellant who is a licensed attorney to pay $1,000 and referred her to Maryland disciplinary authorities.
An appellant appearing without counsel, but licensed to practice in Maryland, used ChatGPT to research and draft her appellate brief. Seven cited cases could not be located. The court affirmed, ordered $1,000 to the appellee, and referred her to Maryland.
Tenth Circuit: A Licensed Lawyer, Appearing Pro Se | TER — Hartzer Press
Clinco v. Commissioner
Event
/ai-in-court/clinco-v-commissioner-fabricated-citations/
Counsel cited four cases in a Tax Court brief and three did not exist. The court called fictitious caselaw a recipe for sanctions but imposed none.
An attorney's post-trial brief cited four cases in support of an argument; the court found three of them to be hallucinations generated by a large language model. It described the filing as a Rule 11(b) violation and imposed no sanction.
Clinco: Three Fabricated Cases in a Tax Brief | The Expert Record — Hartzer Press
Fletcher v. Experian (Fifth Circuit sanction)
Event
/ai-in-court/fletcher-v-experian-fifth-circuit/
In a published opinion the Fifth Circuit sanctioned appellant's counsel $2,500 for an AI-drafted reply brief and for a response that was not forthcoming.
The Fifth Circuit determined that counsel used artificial intelligence to draft a substantial portion of a reply brief without verifying it, and that she was not forthcoming in response to a show-cause order. It imposed a $2,500 sanction.
Fifth Circuit Sanctions Counsel Over an AI Reply Brief | TER — Hartzer Press
Whiting v. City of Athens
Event
/ai-in-court/whiting-v-city-of-athens-sixth-circuit-sanctions/
A federal appeals court found over two dozen fake citations and misstatements across three appeals and imposed fees, double costs, fines and a referral.
The Sixth Circuit found more than two dozen fake citations and misrepresentations in the plaintiff's briefs across three consolidated appeals. After counsel refused to answer its show-cause questions, including whether they used generative AI, the court imposed full appellate fees, double costs, punitive fines and a disciplinary referral; it made no finding on AI use.
Whiting: Sixth Circuit Sanctions for Fake Citations | TER — Hartzer Press
Obi v. Cook County
Event
/ai-in-court/obi-v-cook-county/
N.D. Ill. imposed a $5,000 Rule 11 sanction on a self-represented plaintiff in April 2026 after repeated AI-drafted briefs cited cases that do not exist.
A self-represented plaintiff filed post-judgment briefs, which the court found were generated with AI, containing nonexistent cases and invented quotations after an earlier warning. Chief Judge Virginia M. Kendall struck the briefs and imposed a $5,000 Rule 11 sanction.
Obi v. Cook County: Sanction on a Pro Se Filer | TER — Hartzer Press
In re Prince Global Holdings (Sullivan & Cromwell)
Event
/ai-in-court/prince-global-sullivan-cromwell/
Sullivan & Cromwell told the S.D.N.Y. bankruptcy court in April 2026 that an emergency motion contained AI hallucinations its policies should have caught.
Sullivan & Cromwell filed an emergency motion in a Chapter 15 case that contained inaccurate citations and misquotations, including AI hallucinations. The firm reported the errors to Chief Judge Martin Glenn and filed a corrected motion. No court action has been reported.
Sullivan & Cromwell AI Errors: In re Prince Global | TER — Hartzer Press
Lnu v. Blanche (Ninth Circuit suspension)
Event
/ai-in-court/lnu-v-blanche-ninth-circuit-suspension/
The Ninth Circuit fined two attorneys $2,500 each, suspended them six months, and required AI statements under penalty of perjury for two years.
Two attorneys filed briefs containing fabricated cases and misattributed quotations, then characterized the errors as typographical and denied AI use at argument. The Ninth Circuit fined and suspended both and imposed a two-year certification requirement.
Ninth Circuit Suspends Two Lawyers Over AI Briefs | TER — Hartzer Press
In re Rosslyn2016 (a legal research tool is still a tool)
Event
/ai-in-court/in-re-rosslyn2016-westlaw-precision/
A bankruptcy court rejected the argument that a legal vendor's AI research tool deserves different treatment, finding contempt and ordering $29,877 in fees.
Counsel's motions contained fabricated quotations and invented authority. He denied using ChatGPT and attributed the citations to a legal vendor's AI research feature. The court rejected the distinction, found contempt, ordered CLE and awarded $29,877 in fees.
In re Rosslyn2016: A Legal AI Tool Is Still a Tool | TER — Hartzer Press
Daghra v. Hinkley (government counsel)
Event
/ai-in-court/daghra-v-hinkley-government-lawyer/
A chief district judge found a citation in the Government's filing was likely produced by generative AI and barred non-existent case law in future.
The Government's response to a show-cause order cited a Sixth Circuit case that does not exist. The court found the citation was likely produced by generative AI, declined to impose sanctions, and directed that future filings not include non-existent case law.
Daghra v. Hinkley: The Government's Fake Citation | TER — Hartzer Press
Kleyman Law Group v. Kaloidis
Event
/ai-in-court/kleyman-v-kaloidis-claude-sanctions-referral/
A lawyer cited 23 fictitious decisions across 19 filings after using Claude. A state trial court imposed $10,000, $36,511.35 in fees and a grievance referral.
A lawyer suing on his own firm's fee claim filed 19 submissions citing 23 fictitious decisions, which he testified came from Claude and other generative AI platforms. The court imposed the $10,000 maximum sanction, awarded $36,511.35 in fees and costs, and referred him to a grievance committee.
Kleyman v. Kaloidis: Claude Citations and Sanctions | TER — Hartzer Press
State v. Sandoval (Aarons sanction)
Event
/ai-in-court/sandoval-new-mexico-murder-appeal/
New Mexico Supreme Court held murder-appeal counsel in contempt and ordered $5,000 after his ChatGPT-assisted brief invented witnesses and testimony.
Appellate counsel in a New Mexico murder appeal filed a brief prepared with ChatGPT that described nonexistent witnesses and testimony. The New Mexico Supreme Court held him in contempt, ordered $5,000, removed him and referred him to the Disciplinary Board.
ChatGPT Brief in New Mexico Murder Appeal: Sandoval | TER — Hartzer Press
American Backflow & Fire Prevention, Inc. (NLRB)
Event
/ai-in-court/american-backflow-nlrb-ghost-citations-referral/
The NLRB referred a company's labor relations representative for investigation after her exceptions cited nonexistent record pages and Board decisions.
In American Backflow & Fire Prevention, the company's exceptions to the NLRB cited transcript pages and two Board decisions that do not exist. The Board said AI may have been involved, stated that the verification duty applies to attorneys and non-attorney representatives alike, and referred the representative to its Investigating Officer.
American Backflow: NLRB Refers Ghost Citations | TER — Hartzer Press
SRA v Kumar (Solicitors Disciplinary Tribunal)
Event
/ai-in-court/sra-v-kumar-sdt-ai-strike-off/
The Solicitors Disciplinary Tribunal struck off a foreign lawyer whose AI-drafted filings in his own case cited cases and passages that did not exist.
Defending disciplinary proceedings, a registered foreign lawyer filed an Answer and an email drafted with generative AI that cited nonexistent cases and quotations, and repeated the errors after being warned. The Solicitors Disciplinary Tribunal found the AI allegations proved and struck him off.
SRA v Kumar: First AI Misuse Case Before the SDT | TER — Hartzer Press
Assemblies of God Gospel Church v. Registrar of Societies
Event
/ai-in-court/assemblies-of-god-v-registrar-tanzania-advocate-suspended/
A Tanzanian High Court judge suspended an advocate for six months for citing two nonexistent Court of Appeal decisions and ordered a published apology.
An advocate seeking an extension of time cited and quoted two Court of Appeal decisions that do not exist. The High Court of Tanzania found him guilty of professional misconduct, expunged the cases, suspended him for six months and ordered a published apology in English and Swahili.
Assemblies of God Gospel Church: Advocate Suspended | TER — Hartzer Press
Douglas v. Deutsche Bank National Trust Co.
Event
/ai-in-court/douglas-v-deutsche-bank-brief-struck/
The District of Columbia Court of Appeals struck a bank's appellate brief citing four nonexistent authorities and referred the matter to disciplinary counsel.
Deutsche Bank's brief as appellee cited four authorities that do not exist; the attorney who drafted it said she had used Google's generative AI search tool. The District of Columbia Court of Appeals struck the brief, rejected a proposal to redact the bad citations, and referred the matter to disciplinary counsel.
Douglas v. Deutsche Bank: Appellee's Brief Struck | TER — Hartzer Press
Youssian v. Village of Kings Point
Event
/ai-in-court/youssian-v-kings-point-spellbook-ai/
A New York trial court fined counsel $5,000, awarded $10,000 in fees and struck a petition built on fabricated quotations after Spellbook AI edits.
Counsel's Article 78 petition in Youssian v. Village of Kings Point quoted five real cases with fabricated or altered language after he accepted edits from Spellbook AI in Word. The court fined him $5,000, awarded $10,000 in fees, struck the petition and amended petition, and dismissed.
Youssian: Spellbook AI Quotes and a Struck Petition | TER — Hartzer Press
Virgil v. Experian Information Solutions
Event
/ai-in-court/virgil-v-experian-10000-sanction/
An Indiana federal court personally sanctioned a plaintiff's attorney $10,000 under Rule 11 for five briefs that cited 11 nonexistent cases.
Plaintiff's counsel in Virgil v. Experian filed five briefs containing 11 nonexistent and 4 inaccurate citations. The Southern District of Indiana adopted a recommendation sanctioning him $10,000 personally under Rule 11. He did not cite AI use, and the court said he was responsible either way.
Virgil v. Experian: Rule 11 Sanction for Fake Cases | TER — Hartzer Press
Henry v. Long Island University
Event
/ai-in-court/henry-v-long-island-university-two-year-order/
A federal judge made two lawyers file his opinion in every case they join for two years after a brief relied on fake cases they called typographical errors.
An employment plaintiff's opposition brief cited cases that do not exist and quotations not found in real decisions; counsel called the problems typographical and never filed promised corrections. Because the client had discharged counsel, the court imposed no fine but required two attorneys to file the opinion in every Eastern District of New York case they appear in for two years.
Henry v. LIU: Two Years of Filing the Opinion | The Expert Record — Hartzer Press
Lisandrillo v. Palozzi
Event
/ai-in-court/lisandrillo-v-palozzi-ai-slop/
A Florida appeals court denied a writ petition it called AI slop and ordered counsel to show cause, without the use of AI, why sanctions should not follow.
In Lisandrillo v. Palozzi, Florida's Fourth District denied a 61-page writ petition and described counsel's filings as what appears to be AI slop: real citations, scattershot arguments and allegations contrary to the record. It ordered counsel to show cause, without the use of AI, why sanctions should not be imposed.
Lisandrillo v. Palozzi: AI Slop and a Show-Cause Order | TER — Hartzer Press
Zhejiang Dingli Machinery Co. v. United States
Event
/ai-in-court/zhejiang-dingli-v-united-states-cit-show-cause/
A trade court judge ordered an exporter to support brief statements about a foreign antidumping case or declare who, or what, wrote them. Pending.
A Court of International Trade judge found that a brief's statements that the Commerce Department took part in a Canadian antidumping investigation cited no evidence and appeared to be the product of hallucination by a person or by AI. He ordered the plaintiff to substantiate them or show cause under Rule 11(b)(3), with declarations on who or what wrote the passages; no ruling on the order has been located.
Zhejiang Dingli: Trade Court Asks Who Wrote the Brief | TER — Hartzer Press
Cervantes v. Bianco (Supreme Court of California)
Event
/ai-in-court/cervantes-v-bianco-state-bar-referral/
California's Supreme Court referred counsel to the State Bar after a brief misquoted one of its decisions, an error AI tools apparently contributed to.
In an original writ proceeding over ballots seized by the Riverside County Sheriff, the sheriff's preliminary briefing misquoted and misstated one of the court's own decisions. The firm apologized and withdrew the argument. The Supreme Court of California disapproved the improper use of AI tools and referred the matter to the State Bar.
Cervantes v. Bianco: AI Misquotes and a Bar Referral | TER — Hartzer Press
Aguilar v. The Crawford Group, Inc.
Event
/ai-in-court/aguilar-v-crawford-group-pro-hac-vice-revoked/
A federal judge revoked a lawyer's pro hac vice admission and ordered fees after AI-drafted briefs miscited cases. He had believed his AI tool could not err.
A California lawyer admitted pro hac vice in Massachusetts filed oppositions with fictitious and misattributed citations and said he believed the enterprise version of his AI software did not hallucinate. The court granted Rule 11 sanctions, ordered his firm to pay up to $10,000 in fees and revoked his pro hac vice admission.
Aguilar v. Crawford Group: Pro Hac Vice Revoked | TER — Hartzer Press
Kid International v. City National Bank of Florida
Event
/ai-in-court/kid-international-v-city-national-bank-appellate-fees/
An appellate court affirmed summary judgment and held plaintiffs and counsel jointly liable for fees after an AI-assisted brief cited a nonexistent case.
The plaintiffs' initial brief cited a nonexistent case seven times, attributed fictional quotations to real decisions and misstated holdings; counsel admitted his firm used generative AI and relied on client-supplied research. The appellate court affirmed summary judgment and granted the defendants' motion for appellate attorney's fees as a sanction, with plaintiffs and counsel jointly and severally liable.
Kid International: Fee Sanction for a Fabricated Brief | TER — Hartzer Press
Deepak Raheja v. Omkara Assets Reconstruction
Event
/ai-in-court/raheja-v-omkara-supreme-court-india-ai-compilation/
India's Supreme Court recorded an objection that a party's compilation was full of AI-generated nonexistent judgments; proposed exemplary costs were dropped.
In an insolvency appeal, the respondents objected that the appellants' compilation of judgments was replete with AI-generated nonexistent judgments, and the Advocate-on-Record blamed his clients. The Supreme Court of India proposed exemplary costs but deleted them after the first appellant died, ordering no costs.
Raheja v. Omkara: Fake-Judgment Compilation Rebuked | TER — Hartzer Press
Commonwealth v. Butler (Pa. Super. Ct.)
Event
/ai-in-court/commonwealth-v-butler-pa-superior-false-quotations/
Pennsylvania's Superior Court affirmed a sentence, found counsel falsely quoted six real cases and one that does not exist, and reminded him of his duties.
Defense counsel in Commonwealth v. Butler attributed direct quotations to six real Pennsylvania decisions that do not contain them, and to a seventh case that does not exist. The Superior Court affirmed the sentence, rejected his human error explanation and reminded him of his ethical obligations.
Commonwealth v. Butler: False Quotations on Appeal | TER — Hartzer Press
HHS and NTEU, 74 FLRA No. 87
Event
/ai-in-court/hhs-v-nteu-flra-agency-ai-brief/
The FLRA found 35 of 42 citations in a federal agency's brief fictitious, erroneous or irrelevant, said AI may have been used, and denied the exceptions.
In HHS and NTEU, reviewing a federal agency's exceptions to a telework arbitration award, the Federal Labor Relations Authority found 35 of 42 legal or record citations fictitious, erroneous or irrelevant. It said the agency may have used a generative AI tool, cautioned parties to verify, and denied the exceptions without a sanction.
HHS and NTEU: FLRA Flags Fictitious Agency Citations | TER — Hartzer Press
AI in the Courtroom: AI-generated and altered evidence (20)
Huang v. Tesla
Event
/ai-in-court/huang-v-tesla-musk-deepfake-argument/
Tesla declined to authenticate recordings of Elon Musk, citing deepfakes. A Santa Clara judge called it deeply troubling and ordered a limited deposition.
In a 2023 Autopilot wrongful-death case, Tesla declined to admit recordings of Elon Musk's public statements, noting he is often the subject of deepfakes. The Santa Clara County court ordered a limited 3-hour deposition of Mr. Musk.
Huang v. Tesla: The Musk Deepfake Argument | The Expert Record — Hartzer Press
State of Washington v. Puloka
Event
/ai-in-court/puloka-ai-enhanced-video/
A King County judge excluded a bystander video enhanced with Topaz Labs AI from a 2024 triple-homicide trial after a Frye hearing.
The defense in a Washington triple-homicide case offered a bystander video enhanced with Topaz Labs AI software. After a Frye hearing the King County Superior Court excluded the enhanced version, and the case was tried on the original footage.
State v. Puloka: AI-Enhanced Video Excluded | The Expert Record — Hartzer Press
Maryland v. Darien (Pikesville High deepfake audio)
Event
/ai-in-court/pikesville-principal-deepfake-audio/
A former Pikesville High athletic director entered an Alford plea in 2025 over an AI-generated recording of the principal and received 4 months in jail.
A recording presented as a Baltimore County principal's racist remarks spread in 2024. Police and forensic analysts found it was AI-generated; the school's former athletic director entered an Alford plea and was sentenced to 4 months in jail.
Pikesville High Deepfake Audio Prosecution | The Expert Record — Hartzer Press
Mendones v. Cushman & Wakefield
Event
/ai-in-court/mendones-v-cushman-wakefield-deepfake-exhibits/
An Alameda County judge dismissed a housing case with prejudice in 2025 after finding self-represented plaintiffs filed AI-generated video testimony.
Self-represented plaintiffs in a California housing case filed video "testimony" and altered images in opposing summary judgment. The Alameda County Superior Court found 2 videos were AI-generated and dismissed the case with prejudice.
Mendones v. Cushman & Wakefield: Deepfake Exhibits | TER — Hartzer Press
Florida v. Schinault (ChatGPT image false report)
Event
/ai-in-court/st-petersburg-chatgpt-image-false-report/
A St. Petersburg woman gave police a ChatGPT-generated image of a supposed intruder in 2025. She pleaded no contest to false reporting.
A St. Petersburg, Florida woman gave police a ChatGPT-generated image of a man she said broke in and assaulted her. A detective recognized it from a viral trend; she pleaded no contest to misdemeanor false reporting and received probation and a fine.
St. Petersburg False Report Built on a ChatGPT Image | TER — Hartzer Press
State v. Amyda (deepfake claim without evidence)
Event
/ai-in-court/state-v-amyda-deepfake-authentication/
The Iowa Court of Appeals affirmed admission of a video over a deepfake objection: a speculative claim with no evidence does not defeat authentication.
A defendant argued that a video of the charged abuse may have been a deepfake created with generative AI. The Iowa Court of Appeals affirmed its admission, and on October 9, 2026 the Iowa Supreme Court affirmed as well, finding no identifiable circumstances suggesting a deepfake.
A Deepfake Claim Needs More Than the Claim | The Expert Record — Hartzer Press
Warner v. Gilbarco (ChatGPT work product)
Event
/ai-in-court/warner-v-gilbarco-ai-work-product/
A magistrate judge denied discovery into a self-represented plaintiff's ChatGPT use, holding that using the system did not waive work product protection.
Defendants sought documents about a self-represented plaintiff's use of ChatGPT and argued the use waived work product. The court denied the request, finding the information neither discoverable nor relevant.
Warner v. Gilbarco: No Waiver by ChatGPT Use | The Expert Record — Hartzer Press
Matter of M.S. (M.H.)
Event
/ai-in-court/matter-of-ms-video-authentication/
New York's Court of Appeals reversed abuse findings in 2026, holding 3 hacked security-camera videos were not properly authenticated. Two judges dissented.
A New York family court relied on 3 videos recovered from a hacker's computer to find child abuse. In February 2026 the Court of Appeals, 4-3, held the videos were not properly authenticated and dismissed the petitions, citing the rise of deepfakes.
Matter of M.S.: Video Authentication and Deepfakes | TER — Hartzer Press
United States v. Heppner
Event
/ai-in-court/united-states-v-heppner-ai-privilege/
A Manhattan federal judge ruled that 31 documents a fraud defendant made with Claude, seized by the FBI, were not protected as privileged or work product.
A criminal defendant used Claude, on his own initiative, to analyze his legal exposure and outline defense strategy, and later shared the results with his lawyers. The FBI seized the documents. The court ruled they were not protected by attorney-client privilege or the work product doctrine.
United States v. Heppner: AI Records Not Privileged | TER — Hartzer Press
Burnley v. Valentin
Event
/ai-in-court/burnley-v-valentin-voice-clone-claim/
A federal magistrate judge in Virginia found a recorded call authentic in 2026 after a pro se party said it was an AI clone of his voice.
A pro se litigant accused of breaching a settlement said a recorded phone call was an AI clone of his voice. After a hearing, a federal magistrate judge in Virginia found the recording authentic and enforced the settlement.
Burnley v. Valentin: AI Voice-Clone Claim Rejected | TER — Hartzer Press
Morgan v. V2X (AI work product)
Event
/ai-in-court/morgan-v-v2x-ai-work-product/
A Colorado magistrate judge held a self-represented party's AI use is work product, ordered the tool named, and amended the protective order to restrict AI.
A self-represented plaintiff's use of an AI system was held to fall within work product protection, but he was ordered to name the platform and the protective order was amended to bar confidential material from systems that train on inputs.
Morgan v. V2X: AI Use and Work Product | The Expert Record — Hartzer Press
Tym v. Cerno (AI work product)
Event
/ai-in-court/tym-v-cerno-ai-work-product/
A magistrate judge compelled amended discovery responses but sustained a work product objection over a party's generative AI interactions, requiring a log.
A magistrate judge granted a motion to compel amended discovery responses while sustaining a work product objection over the plaintiff's generative AI interactions and requiring a privilege log.
Tym v. Cerno: AI Sessions and Work Product | The Expert Record — Hartzer Press
Adams v. Anderson (Kentucky Court of Appeals)
Event
/ai-in-court/adams-v-anderson-ai-faked-evidence-contempt/
A parent conceded that testimony and documents were false or altered after AI fabrication was alleged. A 45-day contempt sentence was upheld on appeal.
In a divorce case, a mother testified about a child's urgent-care visit and offered a photo of a prescription bottle that the father said was altered or AI-generated. She admitted lying, pleaded guilty to criminal contempt and served 45 days; the Kentucky Court of Appeals affirmed in an unpublished opinion.
Adams v. Anderson: Jail for AI-Altered Evidence | TER — Hartzer Press
Tate Group Automotive v. Legacy Automotive Capital
Event
/ai-in-court/tate-group-v-legacy-chatgpt-work-product/
A Texas Business Court judge treated a non-lawyer's ChatGPT chats as work product but ordered disclosure of what had been shared with the system.
A non-lawyer party representative's ChatGPT conversations were withheld as work product. After in camera review the court sustained the protection, disagreed with United States v. Heppner, and ordered the party to identify every discovery material it had shared with ChatGPT.
Tate Group: A Party's ChatGPT Chats in Discovery | TER — Hartzer Press
Assini v. Hayward (subpoena to OpenAI quashed)
Event
/ai-in-court/assini-v-hayward-openai-subpoena/
A New York court quashed a subpoena seeking a pro se litigant's OpenAI account records, declining to follow the federal ruling in United States v. Heppner.
Plaintiffs subpoenaed OpenAI for a pro se defendant's account records, including prompts, outputs and uploads. The court quashed the subpoena, holding the contents resemble work product, and expressly declined to follow United States v. Heppner.
A Subpoena to OpenAI for a Litigant's Chats, Quashed | TER — Hartzer Press
Shealy v. Seaside Investments
Event
/ai-in-court/shealy-v-seaside-chatgpt-work-product/
A business court compelled production of ChatGPT prompts and output a party's partner made, holding AI work is not work product without counsel's direction.
A party forwarded loan documents to his partner, who put them through ChatGPT and returned a drafted response. The court ordered the redacted material produced, holding it was not work product because no attorney directed it.
Shealy: ChatGPT Output Ordered Produced | The Expert Record — Hartzer Press
Arbuckle v. Tanner, 2026 NSCA 62
Event
/ai-in-court/arbuckle-v-tanner-ai-transcripts/
An appellant filed AI-made transcripts bearing a court transcriber's certificate. The Nova Scotia Court of Appeal dismissed the appeal with $20,000 costs.
A self-represented appellant filed hearing transcripts made with AI tools that carried a certified transcriber's certificate she had not given. A Nova Scotia Court of Appeal justice dismissed the appeal for failure to perfect it and awarded $20,000 in indemnity costs.
Arbuckle v. Tanner: AI Transcripts, Appeal Dismissed | TER — Hartzer Press
Fry v. Fry (Claude upload and sealing)
Event
/ai-in-court/fry-v-fry-claude-upload-sealing/
A federal judge declined to seal documents a party had uploaded to consumer Claude, and required disclosure of the AI tools, prompts and inputs used.
A self-represented petitioner disclosed uploading redacted filings to non-enterprise Claude. The court granted sealing only in part, reasoning the material had already gone to a private company, and had earlier required a prompt-and-input certification.
Fry v. Fry: A Claude Upload Defeats Sealing | The Expert Record — Hartzer Press
TRAX International Corp. v. United States
Event
/ai-in-court/trax-v-united-states-army-ai-proposal-evaluations/
A federal claims judge ordered an Army AI tool's proposal evaluations into the record, upheld the award, and ordered government counsel to show cause.
In a bid protest, the Court of Federal Claims ordered AI-generated proposal evaluations made by an Army analyst with the FAST TRACK tool added to the administrative record, then upheld the contract award. In a separate order it directed government counsel to show cause why they should not be sanctioned for misrepresenting who had seen those evaluations.
TRAX v. United States: AI Evaluations in a Bid Protest | TER — Hartzer Press
Fontainebleau Florida Hotel v. Botach
Event
/ai-in-court/fontainebleau-v-botach-ai-provider-subpoenas/
A federal magistrate judge let a hotel seek a defamation defendant's AI chats on 3 subjects, rejected a blanket work-product objection and denied consent.
In Fontainebleau v. Botach, a hotel suing for defamation sought the defendant's AI chat histories by document request and by subpoenas to 4 AI providers. A magistrate judge allowed 3 subject categories, barred 3 others, rejected a blanket work-product objection and declined to compel the defendant to consent to disclosure.
Botach: AI Chat Discovery and Provider Subpoenas | TER — Hartzer Press
AI in the Courtroom: Judges, courts and AI (20)
Snell v. United Specialty Insurance (Judge Newsom)
Event
/ai-in-court/newsom-llm-ordinary-meaning-concurrences/
11th Circuit Judge Kevin Newsom queried ChatGPT, Gemini and Claude on ordinary meaning in 2 concurrences in 2024 and set out what he asked and got.
In Snell (May 2024) and Deleon (September 2024), 11th Circuit Judge Kevin Newsom wrote concurrences describing queries he put to large language models about the ordinary meaning of words. The majority decisions did not rest on those queries.
Judge Newsom's Concurrences Using Language Models | TER — Hartzer Press
Colombia: Cartagena ChatGPT ruling and Sentencia T-323/24
Event
/ai-in-court/colombia-chatgpt-tutela-t-323/
A Cartagena judge consulted ChatGPT in a 2023 ruling. In T-323 of 2024 Colombia's Constitutional Court found no due process breach and set principles.
A Cartagena labor judge disclosed that he consulted ChatGPT in a January 2023 ruling on health coverage for a child with autism. Colombia's Constitutional Court reviewed it in 2024, found no due process violation and ordered guidance for the judiciary.
Colombia: Judge's ChatGPT Ruling and Sentencia T-323 | TER — Hartzer Press
Shahid v. Esaam
Event
Georgia Court of Appeals vacated a divorce-case order that relied on 2 fictitious cases and fined the attorney whose name appeared on it $2,500.
A Georgia trial court signed an order, apparently prepared by the husband's attorney, that relied on 2 fictitious cases. The Court of Appeals vacated the order, remanded for a new hearing and imposed a $2,500 penalty on that attorney.
Shahid v. Esaam: Trial Order Citing Fictitious Cases | TER — Hartzer Press
In re CorMedix Securities Litigation (Judge Neals)
Event
/ai-in-court/cormedix-neals-withdrawn-opinion/
A D.N.J. opinion in the CorMedix case was withdrawn after counsel flagged misstated outcomes and invented quotes. The judge cited an intern's ChatGPT use.
Judge Julien Neals withdrew an opinion denying dismissal in the CorMedix securities case after defense counsel identified misstated case outcomes and quotations that did not exist. He later told the Senate that a law school intern had used ChatGPT without authorization.
CorMedix: Judge Neals Withdraws Opinion With Errors | TER — Hartzer Press
Jackson Federation of Teachers v. Fitch (Judge Wingate)
Event
/ai-in-court/wingate-mississippi-dei-order/
A Mississippi federal TRO named nonparties and misquoted the statute. The judge later cited a clerk's use of Perplexity; the 5th Circuit heard argument.
Judge Henry Wingate replaced a temporary restraining order that named parties not in the case and quoted language not in the statute. He later told the Senate a law clerk had used Perplexity. On October 22, 2026 the 5th Circuit vacated his preliminary injunction and directed that the case be reassigned.
Judge Wingate's Mississippi DEI Order and AI Errors | TER — Hartzer Press
Senate Judiciary inquiry into judges' AI use
Event
/ai-in-court/grassley-federal-judges-ai-inquiry/
In October 2025 Sen. Grassley asked 2 federal judges about AI-drafted orders. Both acknowledged staff use; the courts' AO described interim AI guidance.
Senate Judiciary Chairman Chuck Grassley asked Judges Wingate and Neals whether AI was used in drafting error-filled orders. Both acknowledged staff use, and the Administrative Office of the U.S. Courts described its interim AI guidance.
Grassley Inquiry Into Federal Judges' AI Use | The Expert Record — Hartzer Press
Judge Wheeler (Oklahoma) fabricated citations
Event
/ai-in-court/wheeler-oklahoma-judge-fabricated-citations/
An Oklahoma district judge's order contained citations to cases that do not exist. In September 2026 the state Attorney General declined to pursue the claims.
An Oklahoma associate district judge issued an order containing two citations to cases that do not exist, reportedly after using ChatGPT for research. The state Attorney General's office reviewed the claims and determined it would not pursue them further.
An Oklahoma Judge's Order and Two Fake Cases | The Expert Record — Hartzer Press
In re Domestic Partnership of Torres Campos and Munoz
Event
/ai-in-court/torres-campos-v-munoz-temporary-judge-fake-case/
An appellate court held a temporary judge's order erred by relying on a nonexistent case, found the error forfeited, and sanctioned the respondent's counsel.
A temporary judge signed a pet-custody order, drafted by the appellant's counsel, that relied on a nonexistent case first cited by the respondent's counsel. The appellate court held the reliance was an abuse of discretion but found the claim forfeited, affirmed, and sanctioned the respondent's counsel.
Torres Campos: A Signed Order Citing a Fake Case | TER — Hartzer Press
ARIHQ v. Santé Québec, 2026 QCCS 1360
Event
/ai-in-court/arihq-v-sante-quebec-arbitral-award-annulled/
A Canadian superior court annulled an arbitral award whose every doctrinal and case reference was nonexistent, finding the arbitrator's authority delegated.
An arbitral award dismissing a care-home association's payment claim rested entirely on a doctrinal article and decisions that do not exist. The Quebec Superior Court annulled it for failure to follow the arbitral procedure, concluding on the preponderance of evidence that the arbitrator's authority had been delegated.
ARIHQ: Award Resting on Nonexistent Cases Annulled | TER — Hartzer Press
Payne v. State
Event
/ai-in-court/payne-v-state-georgia-new-trial-order/
Georgia Supreme Court vacated a new-trial order drafted by a prosecutor that contained fictitious citations, and suspended her from practice before it.
A Clayton County trial court denied a murder defendant's motion for new trial in an order prepared by the prosecutor, who later acknowledged using AI to draft briefs. The Supreme Court of Georgia vacated the order and suspended her from practice before it for 6 months.
Payne v. State: Prosecutor-Drafted Order Vacated | TER — Hartzer Press
H.C. v. Contreras
Event
/ai-in-court/hc-v-contreras-trial-court-copied-fake-case/
An unpublished appellate opinion reversed a protective-order ruling that reproduced a brief's nonexistent case and misstated the statutory test for abuse.
A trial court's written ruling denying a domestic violence restraining order reproduced, nearly verbatim, a passage from counsel's closing brief that cited a nonexistent case and misstated the statute, after the other side had flagged both errors. The Court of Appeal reversed in an unpublished opinion that does not mention AI.
H.C. v. Contreras: A Ruling Copied From a Brief | TER — Hartzer Press
Henry County Schools v. Grant (party-drafted order)
Event
/ai-in-court/henry-county-schools-v-grant/
The Georgia Court of Appeals vacated a trial court order drafted by counsel that ignored binding precedent, misquoted cases and relied on a hallucinated case.
A trial court adopted an order drafted by plaintiffs' counsel that ignored binding precedent, misquoted cited cases and relied on at least one hallucinated case. The Court of Appeals reversed in part, vacated in part and remanded.
An Order Drafted by Counsel, With a Fake Case In It | TER — Hartzer Press
Pooja Ramesh Singh v. Jammu and Kashmir Bank Ltd.
Event
/ai-in-court/pooja-ramesh-singh-india-nclt-fake-precedents/
India's Supreme Court set aside NCLT and NCLAT orders that relied on 6 fabricated or misquoted precedents the tribunal found through its own research.
India's insolvency tribunal admitted a case relying on precedents that were fabricated or misquoted, found through its own research and repeated on appeal. The Supreme Court set aside both orders on July 2, 2026 and sent the case back.
Pooja Ramesh Singh: Fake Precedents in Tribunal Orders | TER — Hartzer Press
Molawa v. Matjhabeng Local Municipality
Event
/ai-in-court/molawa-v-matjhabeng-disciplinary-chair-citations/
South Africa's Labour Court stayed a disciplinary hearing whose chairperson cited two nonexistent Labour Appeal Court cases, pending review of his rulings.
The chairperson of a municipal disciplinary hearing cited two Labour Appeal Court decisions that do not exist and misdescribed a third authority. The Labour Court stayed the hearing pending review of his rulings, holding that whether the citations came from AI or from the chairperson himself, there was prima facie evidence he had not applied his mind.
Molawa: Hearing Stayed Over Chair's Fake Citations | TER — Hartzer Press
EPFO v. Principal Commissioner of Income Tax
Event
/ai-in-court/epfo-v-pcit-delhi-high-court-nonexistent-judgments/
An Indian High Court set aside a tax appellate order resting on six nonexistent judgments after the retired officer said his Inspector supplied them.
An income tax appellate order relied on six judgments that the Delhi High Court could not find in the law reports. The officer, since retired, said his Inspector had supplied the citations and he had not verified them; the court set the order aside, remanded, and declined to direct disciplinary action because he had retired.
EPFO v. PCIT: Order Citing Fake Judgments Set Aside | TER — Hartzer Press
FJL v TGO (Gauteng Division, Johannesburg)
Event
/ai-in-court/fjl-v-tgo-acting-judge-fictitious-citations/
A South African appeal bench was shown 11 citation errors in an acting judge's judgment, including a nonexistent case. One judge referred it to a regulator.
On a family appeal, a three-judge bench of South Africa's High Court in Johannesburg was shown eleven discrepancies in an acting judge's judgment, including a nonexistent case. The appeal was dismissed on the merits; a concurring judge said AI was the most plausible explanation, made no finding, and forwarded the judgment to the Legal Practice Council.
FJL v TGO: An Acting Judge's Fictitious Citations | TER — Hartzer Press
Phillips v. Parlade (judicial immunity and AI)
Event
/ai-in-court/phillips-v-parlade-judicial-immunity/
A federal court held judicial immunity bars a claim that a state judge relied wholly on AI to issue a ruling, even assuming it were true.
A plaintiff alleged that a state court judge relied wholly on artificial intelligence to issue a ruling without human judgment. The federal court held that absolute judicial immunity bars the claim even assuming the allegation is true, and dismissed with prejudice.
Suing a Judge for Using AI: Immunity Applies | The Expert Record — Hartzer Press
Faiz Enterprise v. State Tax Officer
Event
/ai-in-court/faiz-enterprise-v-state-tax-officer-ai-order/
A tax officer cancelled a GST registration citing AI-generated cases. The High Court quashed the orders and made the department's new AI rules enforceable.
A probationary state tax officer cancelled a business's GST registration in an order citing cases generated by AI that did not exist or did not apply. He admitted the AI use and apologized. The High Court of Gujarat quashed the orders and directed that new departmental instructions on AI be followed on pain of contempt.
Faiz Enterprise: A Tax Officer's AI-Cited Order | TER — Hartzer Press
Vijay Ghanshyam Gadiya v. Union of India
Event
/ai-in-court/gadiya-v-union-of-india-ai-penalty-order/
India's Supreme Court set aside a customs penalty order after checking its authorities itself and finding cases that do not exist or say what was cited.
A customs adjudicating officer imposed a penalty in an order that relied on judgments and articles the appellant said were AI-generated. The Supreme Court of India checked the authorities itself, found some nonexistent and others misapplied, and set aside the order and the High Court judgment upholding it.
Gadiya v. Union of India: Penalty Order Set Aside | TER — Hartzer Press
South African Post Office v Commissioner Fakude
Event
/ai-in-court/sa-post-office-v-fakude-arbitrator-ai-citation/
South Africa's Labour Court stayed a CCMA arbitration after the commissioner's ruling relied on a Constitutional Court judgment that does not exist.
A CCMA commissioner let an unfair dismissal arbitration proceed against the South African Post Office, relying on a Constitutional Court judgment that does not exist. The Labour Court stayed and interdicted the arbitration pending review, inferring that AI was the likely source.
SA Post Office v Fakude: Arbitrator's Fake Citation | TER — Hartzer Press
AI in the Courtroom: AI inside the courtroom (7)
DoNotPay "robot lawyer" and FTC order
Event
/ai-in-court/donotpay-robot-lawyer-ftc/
DoNotPay dropped a 2023 plan to feed AI arguments to a traffic defendant by earpiece. In 2025 the FTC finalized a $193,000 order over its AI lawyer claims.
DoNotPay planned to have a chatbot coach a traffic defendant in court through an earpiece in 2023 and abandoned the plan after warnings from state bars. In 2025 the FTC finalized an order requiring $193,000 and barring unsupported claims that its service could replace a lawyer.
DoNotPay Robot Lawyer Plan and FTC Consent Order | TER — Hartzer Press
Arizona Supreme Court AI news reporters
Event
/ai-in-court/arizona-supreme-court-ai-avatars/
In March 2025 the Arizona Supreme Court began using 2 AI-generated avatars, Daniel and Victoria, to present video summaries of its decisions and news.
The Arizona Supreme Court introduced 2 AI-generated video presenters to announce its rulings and court news. The justices write and approve each script; the software generates the presenter and voice.
Arizona Supreme Court AI-Generated News Reporters | TER — Hartzer Press
Dewald AI avatar argument (N.Y. App. Div., 1st Dept.)
Event
/ai-in-court/dewald-ai-avatar-appellate-division/
A self-represented appellant played an AI-generated avatar to argue his appeal in New York's First Department. The presiding justice stopped the video.
A self-represented appellant presented his oral argument through an AI-generated avatar without telling the court it was synthetic. The presiding justice stopped the video and had him argue in person; the appeal was later decided on the merits.
Dewald AI Avatar Argument, New York Appellate Division | TER — Hartzer Press
State v. Horcasitas (Pelkey AI victim impact statement)
Event
/ai-in-court/pelkey-ai-victim-impact-statement/
A family played an AI-generated video of a slain man at his killer's sentencing in Maricopa County. The judge imposed 10.5 years; the defense appealed.
The family of Christopher Pelkey played an AI-generated video of him delivering a victim impact statement at the sentencing of the man who shot him. Judge Todd Lang praised it and imposed 10.5 years. On September 30, 2026 the Arizona Court of Appeals vacated the sentence, holding the AI video was not reliable and its use was fundamental error.
Pelkey AI Victim Impact Statement, Arizona Sentencing | TER — Hartzer Press
Jones v. Delta Air Lines (ChatGPT at a deposition)
Event
/ai-in-court/jones-v-delta-deposition-chatgpt/
A court instructed a self-represented plaintiff that she could not use ChatGPT or any AI platform to help her answer questions during her deposition.
A self-represented plaintiff had ChatGPT open during her deposition and appeared to be reading from her screen while answering. The court instructed her that she could not use any AI platform to assist her in answering deposition questions.
A Deponent With ChatGPT Open During Testimony | The Expert Record — Hartzer Press
State v. Boles juror ChatGPT research (Paulding County, Ga.)
Event
/ai-in-court/boles-juror-chatgpt-georgia/
A Georgia murder trial was postponed in June 2026 after a prospective juror asked ChatGPT about the case. The judge released the jury pool.
During jury selection in a Paulding County, Georgia murder trial, a prospective juror asked ChatGPT about the case. Judge Dean Bucci released the jury pool, postponed the trial and moved against the juror for contempt.
State v. Boles: Juror Researched Case With ChatGPT | TER — Hartzer Press
Williams v. State (Ind. Ct. App.) AI-assisted transcript
Event
/ai-in-court/williams-indiana-ai-transcript/
The Indiana Court of Appeals noted that a trial transcript's errors suggested generative AI had helped prepare it, and reminded users to proofread output.
In a July 2026 decision, the Indiana Court of Appeals noted that the trial transcript contained misattributions and typos suggesting generative AI had assisted its preparation. It affirmed the conviction and stated that users of such systems must proofread the result.
Williams v. State: AI-Assisted Transcript, Indiana | TER — Hartzer Press
AI in the Courtroom: Rules, orders and guidance (13)
Judge Starr's AI certificate and the Fifth Circuit rule
CreativeWork
/ai-in-court/starr-certification-and-fifth-circuit-ai-rule/
Judge Brantley Starr's 2023 AI certificate requirement in N.D. Texas, and the Fifth Circuit's June 2024 decision not to adopt a similar rule for briefs.
In May 2023 Judge Brantley Starr (N.D. Tex.) required counsel to certify whether generative AI drafted any filing. The Fifth Circuit proposed a similar certificate in November 2023 and decided in June 2024 not to adopt it.
Judge Starr AI Certificate and Fifth Circuit Rule | TER — Hartzer Press
ABA Formal Opinion 512
CreativeWork
/ai-in-court/aba-formal-opinion-512/
What ABA Formal Opinion 512, issued July 29 2024, actually requires of lawyers using generative AI: competence, confidentiality, candor, supervision and fees.
On July 29, 2024 the ABA Standing Committee on Ethics and Professional Responsibility issued Formal Opinion 512, its first opinion applying the Model Rules of Professional Conduct to lawyers' use of generative AI tools.
ABA Formal Opinion 512 (2024), Explained | The Expert Record — Hartzer Press
California Rule of Court 10.430 and Standard 10.80
CreativeWork
/ai-in-court/california-rule-10-430-generative-ai/
California's Judicial Council adopted rule 10.430 and Standard 10.80 in July 2025, requiring courts that allow generative AI to adopt use policies.
On July 18, 2025 the Judicial Council of California adopted rule 10.430 and Standard 10.80. Courts that do not prohibit generative AI by staff and judicial officers had to adopt a use policy by December 15, 2025.
California Rule 10.430 on Generative AI in Courts | TER — Hartzer Press
AI Judicial Guidance, England and Wales (October 2025)
CreativeWork
/ai-in-court/uk-judicial-guidance-on-ai/
England and Wales judiciary issued updated AI guidance for judicial office holders on Oct. 31, 2025, covering confidentiality, accuracy and deepfakes.
On October 31, 2025 the senior judiciary of England and Wales issued updated guidance on AI for judicial office holders and their staff, replacing the April 2025 version. It covers confidentiality, accuracy, AI use by lawyers and litigants, and deepfakes.
England and Wales AI Judicial Guidance (2025) | The Expert Record — Hartzer Press
Civil Justice Council consultation on AI in court documents
CreativeWork
/ai-in-court/civil-justice-council-ai-court-documents/
The Civil Justice Council consulted in 2026 on AI in court documents, including a proposal that expert reports identify and explain any AI used.
In February 2026 the Civil Justice Council of England and Wales consulted on rules for AI use in preparing court documents, including a proposal that expert reports identify and explain AI use. Its mid-2026 update pointed toward transparency for expert evidence.
Civil Justice Council: AI in Court Documents | The Expert Record — Hartzer Press
Proposed Federal Rule of Evidence 707
CreativeWork
/ai-in-court/fre-707-machine-generated-evidence/
Proposed Federal Rule of Evidence 707 on machine-generated evidence: its 2025 text, the public comment, and the May 2026 decision to revise and restudy it.
The federal Advisory Committee on Evidence Rules published proposed Rule 707 for comment in August 2025. In May 2026 it declined to recommend adoption, revised the draft, and set it for further study. It also set aside a deepfake amendment to Rule 901.
Proposed FRE 707: Machine-Generated Evidence | The Expert Record — Hartzer Press
New York courts: interim AI policy and Part 161
CreativeWork
/ai-in-court/new-york-courts-ai-policy-part-161/
New York's court system issued an interim AI policy for judges and staff in October 2025 and adopted Part 161 on AI in court papers effective June 1, 2026.
New York's Unified Court System issued an interim AI policy for judges and staff in October 2025. By an order of March 25, 2026 it adopted Part 161, effective June 1, 2026, which permits AI in preparing court papers without a disclosure requirement.
New York Courts AI Policy and Part 161 Rule | The Expert Record — Hartzer Press
Draft Regulations for Use of AI in Courts, 2026
CreativeWork
/ai-in-court/india-draft-regulations-ai-in-courts-2026/
India's Supreme Court published draft regulations for AI in courts, with absolute prohibitions and a duty to disclose AI help with documents and evidence.
In June 2026 the Supreme Court of India published draft Regulations for Use of Artificial Intelligence in Courts, 2026, for comment. The draft would bar AI-only judicial outcomes and AI credibility scoring, and require parties to disclose AI assistance in documents, pleadings and evidence.
India's Draft Regulations for Use of AI in Courts | TER — Hartzer Press
Connecticut Practice Book Section 4-9 (Generative AI)
CreativeWork
/ai-in-court/connecticut-practice-book-4-9-generative-ai/
Connecticut's Superior Court requires anyone filing a document made with generative AI to verify every citation, authority and item of evidence it produced.
On June 11, 2026 the judges of the Connecticut Superior Court adopted Practice Book Section 4-9, effective June 23, 2026. Anyone who uses generative AI to create or edit a filed document must independently verify all citations, legal authorities and evidence it produced, on pain of sanctions up to nonsuit or default.
Connecticut Practice Book 4-9 on Generative AI | TER — Hartzer Press
Cherokee County Superior Court AI standing order
CreativeWork
/ai-in-court/cherokee-county-ai-standing-order/
Four Superior Court judges ordered anyone using AI on a filing to name the tool and certify under penalty of perjury that every citation exists as cited.
All four judges of the Superior Court of Cherokee County, Georgia, signed a standing order requiring any attorney or self-represented party who uses AI in any capacity on a filing to name the tool and certify, under penalty of perjury, that the document was independently reviewed and every citation verified.
Cherokee County Superior Court AI Standing Order | TER — Hartzer Press
AI clauses in protective orders (Jeffries; Dunn)
CreativeWork
/ai-in-court/ai-protective-orders-jeffries-dunn/
Two 2026 federal orders wrote AI restrictions into protective orders, barring open AI tools for discovery material and limiting which platforms parties use.
In Kansas a magistrate judge extended AI restrictions from confidential material to all discovery materials, barring open AI tools. In Colorado a magistrate judge entered an agreed AI provision but refused to require consent before each new tool.
Protective Orders Start Regulating AI | The Expert Record — Hartzer Press
Tenth Circuit proposed Rule 46.5(B)(4) on generative AI
CreativeWork
/ai-in-court/tenth-circuit-proposed-rule-46-5-generative-ai/
The Tenth Circuit proposed that any filing made with generative AI be checked by a human who verified each authority and quotation. Comments close Oct. 18.
On September 18, 2026 the Tenth Circuit published proposed revisions to its rules for 2027, including a new certification in Rule 46.5(B)(4) that any document prepared with generative AI was reviewed by a human who verified its authorities, quotations and analysis, with sanctions for violations.
Tenth Circuit Proposed Rule 46.5 on Generative AI | TER — Hartzer Press
California SB 574 (Statutes of 2026, chapter 858)
CreativeWork
/ai-in-court/california-sb-574-attorneys-generative-ai/
California enacted SB 574, barring attorneys from delegating the practice of law to generative AI and requiring personal verification of citations in filings.
SB 574, chaptered on September 30, 2026, bars California attorneys from delegating the practice of law to generative AI and requires them to verify and correct AI output and to disclose AI use to the court for documents submitted. It prohibits filing citations the responsible attorney has not personally verified, and covers arbitrators.
California SB 574: Attorneys and Generative AI | TER — Hartzer Press
About and site policies (5)
About this site
WebPage
What this site is for, who maintains it, and the editorial rules it is written under.
What it is for, who maintains it, and the rules it is written under.
About this site | The Expert Record — Hartzer Press
Contact
WebPage
Contact The Expert Record for the working papers, corrections, adoption notices, and questions about the book.
For the working papers, corrections, adoption notices, and questions about the book.
Contact | The Expert Record — Hartzer Press
Privacy
WebPage
How this site handles data.
This site is static HTML. It asks for nothing, and it has no accounts, forms that submit, or advertising. It does count visitors, using Statcounter, and it is delivered through Cloudflare. Both are described below.
Privacy | The Expert Record — Hartzer Press
Terms of Use
WebPage
The terms on which this site is provided.
Provided for information. You may read it, quote it with attribution, and link to it.
Terms of Use | The Expert Record — Hartzer Press
Disclaimer
WebPage
What this site is not.
Nothing on this site is legal advice. It describes conduct and disposition in decided matters and states a professional method. Whether any of it applies to your situation is a question for counsel admitted in your jurisdiction.
Disclaimer | The Expert Record — Hartzer Press