EntityMap

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.

Expert declarations and reports prepared with an AI system, and how tribunals treated the evidence.

Filings that put a nonexistent case, quotation or fact before a court, and what the court did about it.

Deepfakes offered as evidence, AI-enhanced media, and claims that genuine recordings were fabricated.

Orders containing AI errors, judges using AI openly, and the oversight that followed.

Avatars, generated statements, AI in place of counsel, and AI tools used in proceedings.

Rules of evidence, court rules, standing orders, ethics opinions and expert-facing guidance on AI.

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.

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.

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.

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.

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.

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.

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.

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.

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.

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.

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.

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.

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.

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.

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.

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.

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.

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.

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.

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.

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.

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.

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.

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.

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.

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.

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.

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.

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.

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.

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.

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.

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.

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.

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.

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.

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.

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.

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.

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.

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.

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.

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.

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.

A defendant argued that a video may have been a deepfake created with generative AI. The Iowa Court of Appeals affirmed its admission, holding that circumstantial authentication was not defeated by a speculative claim with no evidentiary support.

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.

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.

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.

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.

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.

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.

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.

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.

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 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.

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.

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.

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.

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.

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.

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. Mississippi asked the 5th Circuit to reassign the case.

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.

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.

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.

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.

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.

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.

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.

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.

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.

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.

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.

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.

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.

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.

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.

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.

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.

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.

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.

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.

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.

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.

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.

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.

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.

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.

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.

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.

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.