{ "dataset": "SafeLegalAI — AI-generated evidence and deepfakes in court", "license": "CC BY 4.0 for the coding; US court text public domain; other judgments quoted ≤25 words and linked", "notice": "Provided as is, without warranty of any kind. Not legal advice; no lawyer–client relationship. SafeLegalAI (published by Cognesio LLP) records findings made by courts, regulators and vendors' own published pages; the linked official documents are the record. Third-party names and marks identify their owners' products or bodies and imply no affiliation or endorsement. Terms: https://safelegalai.com/disclaimer No alleged media is hosted or linked.", "url": "https://safelegalai.com/courts/evidence", "repository": "https://github.com/SafeLegalAI/ai-evidence-in-court", "huggingface": "https://huggingface.co/datasets/safelegalaidata/ai-evidence-in-court", "generated": "2026-09-08T07:50:57.757Z", "version": "0.1.1", "decisions": [ { "decision_id": "au-atmo-2026-dungud-v-peachbulk", "case_name": "DunGud Pty Ltd v Peachbulk Pty Ltd", "court": "Australia Trade Marks Office", "country": "AU", "jurisdiction": "australia", "state": null, "date": "2026-08-14", "citation": "[2026] ATMO 154", "docket_number": null, "evidence_type": "chat-output", "ai_role": "acknowledged-ai-output", "offered_by": "applicant", "challenge": "weight", "ruling": "no-weight", "rules_cited": [ "Trade Marks Act 1995 (Cth)" ], "court_finding_quote": "Responses to very particularly worded questions, produced by AI chatbots, are not in and of themselves evidence of a fact.", "text_licence": "quoted-short", "summary": "A party offered ChatGPT and Gemini responses on product classification and markets in a trademark opposition. The hearing officer rejected treating chatbot outputs like dictionary or encyclopaedic evidence because sources, relevance to Australia, currency, and accuracy could not be interrogated.", "sensitive": false, "publication_ban_checked": false, "related_opinion_id": null, "source_url": "https://www.austlii.edu.au/cgi-bin/viewdoc/au/cases/cth/ATMO/2026/154.html?context=1;query=AI%20OR%20%22Artificial%20intelligence%22;mask_path=", "archive_url": null, "fetched_at": "2026-09-07T00:00:00Z", "verification": "fetched-and-read", "lead_source": [ "Damien Charlotin acknowledged-AI-output lead CSV (CC0)" ], "notes": "AustLII live page blocked automated reading; quote and coding came from the lead's copied decision text. Publication-ban status not confirmed.", "url": "https://safelegalai.com/courts/evidence/au-atmo-2026-dungud-v-peachbulk", "notice": "Provided as is, without warranty. Not legal advice. SafeLegalAI (Cognesio LLP) records what courts, regulators, legislatures and vendors' own public pages state; the linked official documents are the record. Names and marks belong to their owners. Anyone named may reply: https://safelegalai.com/report. Full terms: https://safelegalai.com/disclaimer", "page": "https://safelegalai.com/courts/evidence/au-atmo-2026-dungud-v-peachbulk" }, { "decision_id": "au-wasc-2026-perth-day-hospital-v-fitness-cartel", "case_name": "Perth Day Hospital v. Fitness Cartel", "court": "Supreme Court of Western Australia", "country": "AU", "jurisdiction": "australia", "state": "WA", "date": "2026-08-04", "citation": "[2026] WASC 314", "docket_number": null, "evidence_type": "image", "ai_role": "acknowledged-ai-output", "offered_by": "defendant", "challenge": "weight", "ruling": "no-weight", "rules_cited": [], "court_finding_quote": "The court can give very little weight to these documents as being accurate representations of the flooring work.", "text_licence": "quoted-short", "summary": "Defendants used AI-generated images to illustrate original flooring and remedial work. The court noted the AI origin and gave the images very little weight as accurate representations of completed work, resolving the dispute on other evidence rather than generated visuals.", "sensitive": false, "publication_ban_checked": false, "related_opinion_id": null, "source_url": "https://www.austlii.edu.au/cgi-bin/viewdoc/au/cases/wa/WASC/2026/314.html?context=1;query=AI%20OR%20%22Artificial%20intelligence%22;mask_path=", "archive_url": null, "fetched_at": "2026-09-07T00:00:00Z", "verification": "fetched-and-read", "lead_source": [ "Damien Charlotin deepfakes lead CSV (CC0)", "Damien Charlotin acknowledged-AI-output lead CSV (CC0)", "User brief named-case list" ], "notes": "AustLII live page blocked automated reading; quote and coding came from copied lead text. Publication-ban status not confirmed.", "url": "https://safelegalai.com/courts/evidence/au-wasc-2026-perth-day-hospital-v-fitness-cartel", "notice": "Provided as is, without warranty. Not legal advice. SafeLegalAI (Cognesio LLP) records what courts, regulators, legislatures and vendors' own public pages state; the linked official documents are the record. Names and marks belong to their owners. Anyone named may reply: https://safelegalai.com/report. Full terms: https://safelegalai.com/disclaimer", "page": "https://safelegalai.com/courts/evidence/au-wasc-2026-perth-day-hospital-v-fitness-cartel" }, { "decision_id": "us-ca3-2026-bryan-city-philadelphia", "case_name": "Bryan v. City of Philadelphia", "court": "U.S. Court of Appeals for the Third Circuit", "country": "US", "jurisdiction": "us-federal", "state": "PA", "date": "2026-07-17", "citation": null, "docket_number": "No. 26-1313", "evidence_type": "video", "ai_role": "alleged-deepfake", "offered_by": "defendant", "challenge": "authenticity", "ruling": "not-adjudicated", "rules_cited": [ "Heck v. Humphrey", "42 U.S.C. § 1983" ], "court_finding_quote": "Bryan alleged that defendants altered footage played during his criminal proceedings, but the court affirmed dismissal on procedural grounds without deciding authenticity.", "text_licence": "public-domain", "summary": "Bryan alleged that Philadelphia officers manipulated body-camera and dash-camera footage using artificial intelligence and removed scenes from footage played in court. The Third Circuit affirmed dismissal on limitations, Heck, and related procedural grounds, so the alleged AI manipulation was not authenticated or adjudicated on the merits.", "sensitive": false, "publication_ban_checked": true, "related_opinion_id": "ca3-2026-bryan-v-city-of-philadelphia", "source_url": "https://www2.ca3.uscourts.gov/opinarch/261313np.pdf", "archive_url": "https://web.archive.org/web/20260907220232/https://www2.ca3.uscourts.gov/opinarch/261313np.pdf", "fetched_at": "2026-09-07T00:00:00Z", "verification": "fetched-and-read", "lead_source": [ "Damien Charlotin deepfakes lead CSV (CC0)", "User brief named-case list" ], "notes": null, "url": "https://safelegalai.com/courts/evidence/us-ca3-2026-bryan-city-philadelphia", "notice": "Provided as is, without warranty. Not legal advice. SafeLegalAI (Cognesio LLP) records what courts, regulators, legislatures and vendors' own public pages state; the linked official documents are the record. Names and marks belong to their owners. Anyone named may reply: https://safelegalai.com/report. Full terms: https://safelegalai.com/disclaimer", "page": "https://safelegalai.com/courts/evidence/us-ca3-2026-bryan-city-philadelphia" }, { "decision_id": "ca-abkb-2026-r-v-flett", "case_name": "R v Flett", "court": "Court of King's Bench of Alberta", "country": "CA", "jurisdiction": "canada", "state": "AB", "date": "2026-04-17", "citation": "2026 ABKB 296", "docket_number": "240291385Q1", "evidence_type": "video", "ai_role": "alleged-deepfake", "offered_by": "prosecution", "challenge": "authenticity", "ruling": "admitted", "rules_cited": [ "Canada Evidence Act s.31.1" ], "court_finding_quote": "The defence deepfake argument was speculative; circumstantial evidence and police screen recordings supported threshold authenticity.", "text_licence": "quoted-short", "summary": "The Crown tendered three Snapchat story videos showing men in a vehicle and a handgun. Defence counsel invoked AI and deepfakes, but the court relied on contemporaneous police viewing, screen recordings, timestamps, matched occupants, and a seized gun to admit the clips and treat manipulation concerns as weight issues.", "sensitive": false, "publication_ban_checked": false, "related_opinion_id": null, "source_url": "https://www.canlii.org/en/ab/abkb/doc/2026/2026abkb296/2026abkb296.html", "archive_url": null, "fetched_at": "2026-09-07T00:00:00Z", "verification": "fetched-and-read", "lead_source": [ "Damien Charlotin deepfakes lead CSV (CC0)", "User brief named-case list" ], "notes": "CanLII source page could not be fetched by script; text was read from the downloaded lead PDF copy, so publication-ban header was not independently confirmed.", "url": "https://safelegalai.com/courts/evidence/ca-abkb-2026-r-v-flett", "notice": "Provided as is, without warranty. Not legal advice. SafeLegalAI (Cognesio LLP) records what courts, regulators, legislatures and vendors' own public pages state; the linked official documents are the record. Names and marks belong to their owners. Anyone named may reply: https://safelegalai.com/report. Full terms: https://safelegalai.com/disclaimer", "page": "https://safelegalai.com/courts/evidence/ca-abkb-2026-r-v-flett" }, { "decision_id": "fr-tj-clermont-2026-no-25-04051", "case_name": "N° 25/04051", "court": "Tribunal judiciaire de Clermont-Ferrand", "country": "FR", "jurisdiction": "other", "state": null, "date": "2026-04-16", "citation": null, "docket_number": "25/04051", "evidence_type": "expert-report", "ai_role": "acknowledged-ai-output", "offered_by": "plaintiff", "challenge": "weight", "ruling": "not-adjudicated", "rules_cited": [], "court_finding_quote": null, "text_licence": "quoted-short", "summary": "Counsel offered an AI-assisted damages evaluation, but the issue became moot or unnecessary to the disposition on the available lead. No public primary court copy was found; the row preserves the lead for future verification against the tribunal decision or a publisher copy.", "sensitive": false, "publication_ban_checked": false, "related_opinion_id": null, "source_url": "https://www.doctrine.fr/d/TJ/Clermont-Ferrand/2026/TJP3F1C429C80CD0EC0F4D0", "archive_url": "https://web.archive.org/web/20260907221039/https://www.doctrine.fr/restriction?redirect_to=%2Fd%2FTJ%2FClermont-Ferrand%2F2026%2FTJP3F1C429C80CD0EC0F4D0&require_login=false&sourcePage=Decision&kind=decisions", "fetched_at": "2026-09-07T00:00:00Z", "verification": "link-only", "lead_source": [ "Damien Charlotin acknowledged-AI-output lead CSV (CC0)" ], "notes": "Paywalled Doctrine lead only; no primary copy located and no publication-ban check confirmed.", "url": "https://safelegalai.com/courts/evidence/fr-tj-clermont-2026-no-25-04051", "notice": "Provided as is, without warranty. Not legal advice. SafeLegalAI (Cognesio LLP) records what courts, regulators, legislatures and vendors' own public pages state; the linked official documents are the record. Names and marks belong to their owners. Anyone named may reply: https://safelegalai.com/report. Full terms: https://safelegalai.com/disclaimer", "page": "https://safelegalai.com/courts/evidence/fr-tj-clermont-2026-no-25-04051" }, { "decision_id": "ca-qctat-2026-cote-placements-jf-desgagne", "case_name": "Côté et Placements J.F. Desgagné inc.", "court": "Tribunal administratif du travail", "country": "CA", "jurisdiction": "canada", "state": "QC", "date": "2026-03-26", "citation": "2026 QCTAT 1311", "docket_number": null, "evidence_type": "chat-output", "ai_role": "acknowledged-ai-output", "offered_by": "applicant", "challenge": "weight", "ruling": "excluded", "rules_cited": [ "Quebec administrative evidence principles" ], "court_finding_quote": "Le Tribunal n'accorde aucune valeur probante aux réponses générées par l'IA.", "text_licence": "quoted-short", "summary": "A self-represented worker filed voluminous Google AI material on medical and legal issues. The Tribunal found the outputs unreliable because sources were not tied to answers, authors and expertise were unknown, and the material lacked contextualization, excluding it and relying on medical records and testimony.", "sensitive": true, "publication_ban_checked": false, "related_opinion_id": null, "source_url": "https://www.canlii.org/fr/qc/qctat/doc/2026/2026qctat1311/2026qctat1311.html", "archive_url": null, "fetched_at": "2026-09-07T00:00:00Z", "verification": "fetched-and-read", "lead_source": [ "Damien Charlotin acknowledged-AI-output lead CSV (CC0)" ], "notes": "CanLII page could not be independently read by script; downloaded lead PDF copy was read.", "url": "https://safelegalai.com/courts/evidence/ca-qctat-2026-cote-placements-jf-desgagne", "notice": "Provided as is, without warranty. Not legal advice. SafeLegalAI (Cognesio LLP) records what courts, regulators, legislatures and vendors' own public pages state; the linked official documents are the record. Names and marks belong to their owners. Anyone named may reply: https://safelegalai.com/report. Full terms: https://safelegalai.com/disclaimer", "page": "https://safelegalai.com/courts/evidence/ca-qctat-2026-cote-placements-jf-desgagne" }, { "decision_id": "us-edva-2026-burnley-v-valentin", "case_name": "Burnley v. Valentin", "court": "U.S. District Court for the Eastern District of Virginia", "country": "US", "jurisdiction": "us-federal", "state": "VA", "date": "2026-03-18", "citation": null, "docket_number": "3:23-cv-00160", "evidence_type": "audio", "ai_role": "voice-clone", "offered_by": "defendant", "challenge": "authenticity", "ruling": "admitted", "rules_cited": [ "Fed. R. Evid. 901(a)", "Fed. R. Evid. 901(b)(5)" ], "court_finding_quote": "Federal Rule of Evidence 901(a) requires only that the proponent of evidence produce evidence sufficient to support a finding that the item is what the proponent claims it is.", "text_licence": "public-domain", "summary": "Walburn offered a phone recording to enforce a settlement. Burnley alleged the voice was either someone else or an AI clone. The district court accepted sworn declarations about creation and transfer of the recording, plus voice identification under Rule 901(b)(5), and enforced the settlement with fees.", "sensitive": false, "publication_ban_checked": true, "related_opinion_id": "edva-2026-burnley-v-valentin", "source_url": "https://www.damiencharlotin.com/documents/2966/0.pdf", "archive_url": null, "fetched_at": "2026-09-07T00:00:00Z", "verification": "fetched-and-read", "lead_source": [ "Damien Charlotin deepfakes lead CSV (CC0)" ], "notes": null, "url": "https://safelegalai.com/courts/evidence/us-edva-2026-burnley-v-valentin", "notice": "Provided as is, without warranty. Not legal advice. SafeLegalAI (Cognesio LLP) records what courts, regulators, legislatures and vendors' own public pages state; the linked official documents are the record. Names and marks belong to their owners. Anyone named may reply: https://safelegalai.com/report. Full terms: https://safelegalai.com/disclaimer", "page": "https://safelegalai.com/courts/evidence/us-edva-2026-burnley-v-valentin" }, { "decision_id": "eu-euipo-2026-exagres-v-gres-de-aragon", "case_name": "Exagres v. Gres de Aragon", "court": "European Union Intellectual Property Office", "country": "ES", "jurisdiction": "eu", "state": null, "date": "2026-02-19", "citation": null, "docket_number": null, "evidence_type": "chat-output", "ai_role": "acknowledged-ai-output", "offered_by": "applicant", "challenge": "weight", "ruling": "no-weight", "rules_cited": [ "EU trade mark opposition practice" ], "court_finding_quote": "ChatGPT vocabulary analysis was not accepted as probative evidence of consumer understanding.", "text_licence": "quoted-short", "summary": "A party submitted ChatGPT vocabulary analysis in an EUIPO dispute. The tribunal ignored or gave no weight to the chatbot material because it did not provide verifiable evidence of market or consumer understanding and could not substitute for admissible proof.", "sensitive": false, "publication_ban_checked": false, "related_opinion_id": null, "source_url": "https://www.damiencharlotin.com/documents/1686/EUIPODBAC1240C9C6C4F78545.pdf", "archive_url": "https://web.archive.org/web/20260817232955/https://websitedc.s3.amazonaws.com/documents/EUIPODBAC1240C9C6C4F78545.pdf", "fetched_at": "2026-09-07T00:00:00Z", "verification": "fetched-and-read", "lead_source": [ "Damien Charlotin acknowledged-AI-output lead CSV (CC0)" ], "notes": null, "url": "https://safelegalai.com/courts/evidence/eu-euipo-2026-exagres-v-gres-de-aragon", "notice": "Provided as is, without warranty. Not legal advice. SafeLegalAI (Cognesio LLP) records what courts, regulators, legislatures and vendors' own public pages state; the linked official documents are the record. Names and marks belong to their owners. Anyone named may reply: https://safelegalai.com/report. Full terms: https://safelegalai.com/disclaimer", "page": "https://safelegalai.com/courts/evidence/eu-euipo-2026-exagres-v-gres-de-aragon" }, { "decision_id": "us-nycoa-2026-matter-ms", "case_name": "Matter of M.S. (M.H.)", "court": "New York Court of Appeals", "country": "US", "jurisdiction": "us-state", "state": "NY", "date": "2026-02-17", "citation": "2026 NY Slip Op 00825", "docket_number": "No. 7", "evidence_type": "video", "ai_role": "alleged-deepfake", "offered_by": "petitioner", "challenge": "authenticity", "ruling": "excluded", "rules_cited": [ "People v. Patterson", "People v. Price", "Family Court Act article 10" ], "court_finding_quote": "Because the foundation offered for the videos was insufficient, we hold that Family Court erred in admitting the videos into evidence and therefore reverse.", "text_licence": "public-domain", "summary": "Erie County relied entirely on hacked home-surveillance videos in child-protection proceedings. The Court of Appeals held the provenance, chain-of-custody gap, and lack of testimony from the camera operator or hacker made the foundation insufficient, excluding the videos and dismissing the petitions.", "sensitive": true, "publication_ban_checked": true, "related_opinion_id": "ny-2026-matter-of-ms", "source_url": "https://www.nycourts.gov/reporter/3dseries/2026/2026_00825.htm", "archive_url": null, "fetched_at": "2026-09-07T00:00:00Z", "verification": "fetched-and-read", "lead_source": [ "Damien Charlotin deepfakes lead CSV (CC0)", "User brief named-case list" ], "notes": null, "url": "https://safelegalai.com/courts/evidence/us-nycoa-2026-matter-ms", "notice": "Provided as is, without warranty. Not legal advice. SafeLegalAI (Cognesio LLP) records what courts, regulators, legislatures and vendors' own public pages state; the linked official documents are the record. Names and marks belong to their owners. Anyone named may reply: https://safelegalai.com/report. Full terms: https://safelegalai.com/disclaimer", "page": "https://safelegalai.com/courts/evidence/us-nycoa-2026-matter-ms" }, { "decision_id": "fr-ca-rouen-2026-no-25-01677", "case_name": "n° 25/01677", "court": "Cour d'appel de Rouen", "country": "FR", "jurisdiction": "other", "state": null, "date": "2026-02-11", "citation": null, "docket_number": "25/01677", "evidence_type": "expert-report", "ai_role": "acknowledged-ai-output", "offered_by": "plaintiff", "challenge": "weight", "ruling": "no-weight", "rules_cited": [], "court_finding_quote": "The court preferred non-AI expert evidence over AI-assisted assertions about rain patterns.", "text_licence": "quoted-short", "summary": "Counsel offered AI-assisted material about rainfall patterns in a French civil appeal. The court preferred conventional expert or evidentiary material and did not treat the AI output as probative proof. The public court page required JavaScript, so the row remains lightly verified.", "sensitive": false, "publication_ban_checked": false, "related_opinion_id": null, "source_url": "https://www.courdecassation.fr/decision/698d70a5cdc6046d47fd91bb", "archive_url": "https://web.archive.org/web/20260907221150/https://www.courdecassation.fr/decision/698d70a5cdc6046d47fd91bb", "fetched_at": "2026-09-07T00:00:00Z", "verification": "page-read-no-document", "lead_source": [ "Damien Charlotin acknowledged-AI-output lead CSV (CC0)" ], "notes": "Court page showed a JavaScript redirect to the reader; no full primary text was accessible through fetch and publication-ban status not confirmed.", "url": "https://safelegalai.com/courts/evidence/fr-ca-rouen-2026-no-25-01677", "notice": "Provided as is, without warranty. Not legal advice. SafeLegalAI (Cognesio LLP) records what courts, regulators, legislatures and vendors' own public pages state; the linked official documents are the record. Names and marks belong to their owners. Anyone named may reply: https://safelegalai.com/report. Full terms: https://safelegalai.com/disclaimer", "page": "https://safelegalai.com/courts/evidence/fr-ca-rouen-2026-no-25-01677" }, { "decision_id": "ca-bcsc-2026-head-v-doe", "case_name": "Head v. Doe", "court": "Supreme Court of British Columbia", "country": "CA", "jurisdiction": "canada", "state": "BC", "date": "2026-02-04", "citation": "2026 BCSC 184", "docket_number": "M185379", "evidence_type": "video", "ai_role": "alleged-deepfake", "offered_by": "defendant", "challenge": "authenticity", "ruling": "admitted", "rules_cited": [ "Canada Evidence Act s.31.1" ], "court_finding_quote": "The AI tool printout was given no weight; expert metadata and continuity evidence showed the GoPro file was camera-original.", "text_licence": "quoted-short", "summary": "Mr. Head challenged a GoPro accident video with an online AI-tool printout saying deepfake detected. The court accepted witness evidence about the camera and SD card plus forensic analysis of metadata and frame continuity, rejected the unsupported AI-tool output, relied on the video, and dismissed the claim.", "sensitive": false, "publication_ban_checked": false, "related_opinion_id": null, "source_url": "https://www.canlii.org/en/bc/bcsc/doc/2026/2026bcsc184/2026bcsc184.html", "archive_url": null, "fetched_at": "2026-09-07T00:00:00Z", "verification": "fetched-and-read", "lead_source": [ "Damien Charlotin deepfakes lead CSV (CC0)", "User brief named-case list" ], "notes": "CanLII page could not be independently read by script; decision text was read from the downloaded lead PDF copy.", "url": "https://safelegalai.com/courts/evidence/ca-bcsc-2026-head-v-doe", "notice": "Provided as is, without warranty. Not legal advice. SafeLegalAI (Cognesio LLP) records what courts, regulators, legislatures and vendors' own public pages state; the linked official documents are the record. Names and marks belong to their owners. Anyone named may reply: https://safelegalai.com/report. Full terms: https://safelegalai.com/disclaimer", "page": "https://safelegalai.com/courts/evidence/ca-bcsc-2026-head-v-doe" }, { "decision_id": "us-iowa-2026-state-v-amyda", "case_name": "State v. Amyda", "court": "Iowa Court of Appeals", "country": "US", "jurisdiction": "us-state", "state": "IA", "date": "2026-01-28", "citation": null, "docket_number": "No. 24-1563", "evidence_type": "video", "ai_role": "alleged-deepfake", "offered_by": "prosecution", "challenge": "authenticity", "ruling": "admitted", "rules_cited": [ "Iowa R. Evid. 5.901(a)", "Iowa R. Evid. 5.901(b)(4)" ], "court_finding_quote": "Amyda’s speculative claims that the video was a “deepfake” without any evidentiary support did not raise a genuine question about the video’s authenticity.", "text_licence": "public-domain", "summary": "The prosecution introduced a Facebook Messenger video showing sexual contact with a sleeping victim. The court upheld authentication through circumstantial evidence identifying the victim, bedroom, clothing, and the defendant’s tattooed hand, rejecting Amyda’s unsupported deepfake theory and affirming the conviction.", "sensitive": true, "publication_ban_checked": true, "related_opinion_id": "iowa-2026-state-v-amyda", "source_url": "https://www.damiencharlotin.com/documents/2967/2026-24-1563.pdf", "archive_url": null, "fetched_at": "2026-09-07T00:00:00Z", "verification": "fetched-and-read", "lead_source": [ "Damien Charlotin deepfakes lead CSV (CC0)", "User brief named-case list" ], "notes": "Official Iowa PDF endpoint returned 404 during this session; downloaded lead PDF copy was read.", "url": "https://safelegalai.com/courts/evidence/us-iowa-2026-state-v-amyda", "notice": "Provided as is, without warranty. Not legal advice. SafeLegalAI (Cognesio LLP) records what courts, regulators, legislatures and vendors' own public pages state; the linked official documents are the record. Names and marks belong to their owners. Anyone named may reply: https://safelegalai.com/report. Full terms: https://safelegalai.com/disclaimer", "page": "https://safelegalai.com/courts/evidence/us-iowa-2026-state-v-amyda" }, { "decision_id": "us-wdtex-2026-united-states-v-hernandez", "case_name": "United States v. Hernandez", "court": "U.S. District Court for the Western District of Texas", "country": "US", "jurisdiction": "us-federal", "state": "TX", "date": "2026-01-21", "citation": null, "docket_number": "3:25-cr-02384", "evidence_type": "mixed", "ai_role": "ai-generated-submitted", "offered_by": "prosecution", "challenge": "sanctions", "ruling": "sanction", "rules_cited": [ "18 U.S.C. § 1512(c)(2)" ], "court_finding_quote": "Public reports describe a guilty plea after fabricated AI audiovisual evidence allegedly caused an innocent person to be indicted.", "text_licence": "public-domain", "summary": "Public reporting describes a confidential informant pleading guilty to obstruction after providing fabricated AI audiovisual evidence to federal prosecutors, allegedly leading to an innocent person’s indictment. No official plea transcript or factual basis was located, so the row is coded cautiously as link-only.", "sensitive": false, "publication_ban_checked": false, "related_opinion_id": "wdtex-2026-united-states-v-hernandez", "source_url": "https://www.esquiresolutions.com/federal-judges-set-bar-for-deepfake-evidence-challenges/", "archive_url": "https://web.archive.org/web/20260907220814/https://www.esquiresolutions.com/federal-judges-set-bar-for-deepfake-evidence-challenges/", "fetched_at": "2026-09-07T00:00:00Z", "verification": "link-only", "lead_source": [ "User brief named-case list" ], "notes": "Official W.D. Tex. plea or factual-basis document not located; coded from secondary legal reporting and user lead. Publication-ban status not confirmed.", "url": "https://safelegalai.com/courts/evidence/us-wdtex-2026-united-states-v-hernandez", "notice": "Provided as is, without warranty. Not legal advice. SafeLegalAI (Cognesio LLP) records what courts, regulators, legislatures and vendors' own public pages state; the linked official documents are the record. Names and marks belong to their owners. Anyone named may reply: https://safelegalai.com/report. Full terms: https://safelegalai.com/disclaimer", "page": "https://safelegalai.com/courts/evidence/us-wdtex-2026-united-states-v-hernandez" }, { "decision_id": "ca-bccrt-2026-shirran-v-owners-strata-plan-kas-4004", "case_name": "Shirran v. The Owners, Strata Plan KAS 4004", "court": "British Columbia Civil Resolution Tribunal", "country": "CA", "jurisdiction": "canada", "state": "BC", "date": "2026-01-14", "citation": null, "docket_number": null, "evidence_type": "expert-report", "ai_role": "acknowledged-ai-output", "offered_by": "applicant", "challenge": "weight", "ruling": "no-weight", "rules_cited": [], "court_finding_quote": "Expert evidence must come from a known individual with known qualifications and issue-specific knowledge.", "text_licence": "quoted-short", "summary": "Applicants filed several AI-generated opinions about whether plants caused a pest problem. The CRT placed no weight on them because expert evidence must come from an identifiable qualified person with issue-specific knowledge, and treated anonymous or general internet materials similarly.", "sensitive": false, "publication_ban_checked": true, "related_opinion_id": null, "source_url": "https://decisions.civilresolutionbc.ca/crt/sd/en/item/528925/index.do", "archive_url": "https://web.archive.org/web/20260907221239/https://decisions.civilresolutionbc.ca/crt/sd/en/item/528925/index.do", "fetched_at": "2026-09-07T00:00:00Z", "verification": "fetched-and-read", "lead_source": [ "Damien Charlotin acknowledged-AI-output lead CSV (CC0)" ], "notes": null, "url": "https://safelegalai.com/courts/evidence/ca-bccrt-2026-shirran-v-owners-strata-plan-kas-4004", "notice": "Provided as is, without warranty. Not legal advice. SafeLegalAI (Cognesio LLP) records what courts, regulators, legislatures and vendors' own public pages state; the linked official documents are the record. Names and marks belong to their owners. Anyone named may reply: https://safelegalai.com/report. Full terms: https://safelegalai.com/disclaimer", "page": "https://safelegalai.com/courts/evidence/ca-bccrt-2026-shirran-v-owners-strata-plan-kas-4004" }, { "decision_id": "it-modena-2025-whatsapp-ai-contest", "case_name": "Tribunale di Modena, 29 dicembre 2025", "court": "Tribunale di Modena", "country": "IT", "jurisdiction": "other", "state": null, "date": "2025-12-29", "citation": null, "docket_number": null, "evidence_type": "screenshot", "ai_role": "alleged-deepfake", "offered_by": "plaintiff", "challenge": "authenticity", "ruling": "admitted", "rules_cited": [ "Codice civile art. 2712" ], "court_finding_quote": "La contestazione datoriale era oltremodo generica e priva di una specifica perizia.", "text_licence": "quoted-short", "summary": "An employer argued that WhatsApp screenshots offered by a worker had been manipulated or created with AI. The Modena labour court treated the challenge as generic under Italian evidence rules, found no expert or specific inconsistencies, admitted the documents, and awarded reinstatement and damages.", "sensitive": false, "publication_ban_checked": false, "related_opinion_id": null, "source_url": "https://ntplusdiritto.ilsole24ore.com/art/il-paradosso-deepfake-perfetto-quando-l-ia-e-troppo-brava-essere-contestata-AITMx84", "archive_url": "http://web.archive.org/web/20260518212836/https://ntplusdiritto.ilsole24ore.com/art/il-paradosso-deepfake-perfetto-quando-l-ia-e-troppo-brava-essere-contestata-AITMx84", "fetched_at": "2026-09-07T00:00:00Z", "verification": "link-only", "lead_source": [ "Damien Charlotin deepfakes lead CSV (CC0)", "User brief named-case list" ], "notes": "Decision text was not found on an official court site; coded from a legal-news report and lead copy. Publication-ban status not confirmed.", "url": "https://safelegalai.com/courts/evidence/it-modena-2025-whatsapp-ai-contest", "notice": "Provided as is, without warranty. Not legal advice. SafeLegalAI (Cognesio LLP) records what courts, regulators, legislatures and vendors' own public pages state; the linked official documents are the record. Names and marks belong to their owners. Anyone named may reply: https://safelegalai.com/report. Full terms: https://safelegalai.com/disclaimer", "page": "https://safelegalai.com/courts/evidence/it-modena-2025-whatsapp-ai-contest" }, { "decision_id": "ca-oncj-2025-r-v-medow", "case_name": "R. v. Medow", "court": "Ontario Court of Justice", "country": "CA", "jurisdiction": "canada", "state": "ON", "date": "2025-12-11", "citation": "2025 ONCJ 661", "docket_number": null, "evidence_type": "video", "ai_role": "alleged-deepfake", "offered_by": "prosecution", "challenge": "authenticity", "ruling": "admitted", "rules_cited": [ "Canada Evidence Act s.31.1" ], "court_finding_quote": "Officer identification and in-court review supported admission despite judicial notice that deepfake technology exists.", "text_licence": "quoted-short", "summary": "The accused alleged police body-worn and in-car camera videos were digitally altered deepfakes. The court took notice of deepfake technology but accepted officer testimony that the footage accurately depicted events, found no evidentiary support for manipulation, admitted the recordings, and convicted the accused.", "sensitive": false, "publication_ban_checked": false, "related_opinion_id": null, "source_url": "https://www.canlii.org/en/on/oncj/doc/2025/2025oncj661/2025oncj661.html", "archive_url": null, "fetched_at": "2026-09-07T00:00:00Z", "verification": "fetched-and-read", "lead_source": [ "Damien Charlotin deepfakes lead CSV (CC0)", "User brief named-case list" ], "notes": "CanLII page could not be independently read by script; text was read from the downloaded lead PDF copy.", "url": "https://safelegalai.com/courts/evidence/ca-oncj-2025-r-v-medow", "notice": "Provided as is, without warranty. Not legal advice. SafeLegalAI (Cognesio LLP) records what courts, regulators, legislatures and vendors' own public pages state; the linked official documents are the record. Names and marks belong to their owners. Anyone named may reply: https://safelegalai.com/report. Full terms: https://safelegalai.com/disclaimer", "page": "https://safelegalai.com/courts/evidence/ca-oncj-2025-r-v-medow" }, { "decision_id": "il-jerusalem-labor-2025-biton-v-nii", "case_name": "Biton v. National Insurance Institute", "court": "Jerusalem Regional Labor Court", "country": "IL", "jurisdiction": "israel", "state": null, "date": "2025-10-16", "citation": null, "docket_number": null, "evidence_type": "expert-report", "ai_role": "acknowledged-ai-output", "offered_by": "plaintiff", "challenge": "expert-required", "ruling": "other", "rules_cited": [], "court_finding_quote": "The court permitted submission of underlying studies while cautioning that ChatGPT should be used carefully.", "text_licence": "quoted-short", "summary": "Plaintiff’s counsel used ChatGPT to argue that pulling a heavy cart could injure the back and to draft questions for a medical expert. The court rejected fact-specific AI-derived questions but allowed underlying medical studies allegedly cited by ChatGPT to be submitted for expert response.", "sensitive": true, "publication_ban_checked": false, "related_opinion_id": null, "source_url": "https://www.damiencharlotin.com/documents/886/biton35.pdf", "archive_url": null, "fetched_at": "2026-09-07T00:00:00Z", "verification": "fetched-and-read", "lead_source": [ "Damien Charlotin acknowledged-AI-output lead CSV (CC0)" ], "notes": "Hebrew PDF copy read; no official publication-ban check confirmed.", "url": "https://safelegalai.com/courts/evidence/il-jerusalem-labor-2025-biton-v-nii", "notice": "Provided as is, without warranty. Not legal advice. SafeLegalAI (Cognesio LLP) records what courts, regulators, legislatures and vendors' own public pages state; the linked official documents are the record. Names and marks belong to their owners. Anyone named may reply: https://safelegalai.com/report. Full terms: https://safelegalai.com/disclaimer", "page": "https://safelegalai.com/courts/evidence/il-jerusalem-labor-2025-biton-v-nii" }, { "decision_id": "us-wdwa-2025-kirkham-v-bellingham", "case_name": "Kirkham v. City of Bellingham", "court": "U.S. District Court for the Western District of Washington", "country": "US", "jurisdiction": "us-federal", "state": "WA", "date": "2025-10-09", "citation": null, "docket_number": "C24-1625-DGE-SKV", "evidence_type": "video", "ai_role": "alleged-deepfake", "offered_by": "defendant", "challenge": "expert-required", "ruling": "not-adjudicated", "rules_cited": [ "Fed. R. Evid. 706" ], "court_finding_quote": "Plaintiff lacks any evidence other than his own assertions that evidence tampering has occurred.", "text_licence": "public-domain", "summary": "Kirkham alleged officers used AI to replace body-worn-camera footage with deepfake video and sought a neutral expert. The court denied counsel and expert appointment because the claims rested only on his assertions, Rule 706 cannot build a party’s case, and the request was premature.", "sensitive": false, "publication_ban_checked": true, "related_opinion_id": "wdwash-2025-kirkham-v-bellingham", "source_url": "https://www.damiencharlotin.com/documents/2964/0_2.pdf", "archive_url": null, "fetched_at": "2026-09-07T00:00:00Z", "verification": "fetched-and-read", "lead_source": [ "Damien Charlotin deepfakes lead CSV (CC0)", "User brief named-case list" ], "notes": null, "url": "https://safelegalai.com/courts/evidence/us-wdwa-2025-kirkham-v-bellingham", "notice": "Provided as is, without warranty. Not legal advice. SafeLegalAI (Cognesio LLP) records what courts, regulators, legislatures and vendors' own public pages state; the linked official documents are the record. Names and marks belong to their owners. Anyone named may reply: https://safelegalai.com/report. Full terms: https://safelegalai.com/disclaimer", "page": "https://safelegalai.com/courts/evidence/us-wdwa-2025-kirkham-v-bellingham" }, { "decision_id": "us-dnev-2025-demissie-v-ford", "case_name": "Demissie v. Ford", "court": "U.S. District Court for the District of Nevada", "country": "US", "jurisdiction": "us-federal", "state": "NV", "date": "2025-10-02", "citation": null, "docket_number": "2:25-cv-00504", "evidence_type": "video", "ai_role": "alleged-deepfake", "offered_by": "defendant", "challenge": "sanctions", "ruling": "not-adjudicated", "rules_cited": [ "Fed. R. Civ. P. 11", "Fed. R. Civ. P. 37" ], "court_finding_quote": "The Court found no credible evidence supporting plaintiff's claim that defense counsel fabricated or altered body-worn-camera footage.", "text_licence": "public-domain", "summary": "A pro se plaintiff accused defendants and counsel of altering or fabricating body-worn-camera exhibits with AI and sought sanctions. The court reviewed the record and sealed exhibits, found no credible support for the accusation, denied relief, and left reconsideration possible if real evidence emerged.", "sensitive": false, "publication_ban_checked": true, "related_opinion_id": "dnev-2025-demissie-v-ford", "source_url": "https://www.damiencharlotin.com/documents/2965/0_1.pdf", "archive_url": null, "fetched_at": "2026-09-07T00:00:00Z", "verification": "fetched-and-read", "lead_source": [ "Damien Charlotin deepfakes lead CSV (CC0)", "User brief named-case list" ], "notes": null, "url": "https://safelegalai.com/courts/evidence/us-dnev-2025-demissie-v-ford", "notice": "Provided as is, without warranty. Not legal advice. SafeLegalAI (Cognesio LLP) records what courts, regulators, legislatures and vendors' own public pages state; the linked official documents are the record. Names and marks belong to their owners. Anyone named may reply: https://safelegalai.com/report. Full terms: https://safelegalai.com/disclaimer", "page": "https://safelegalai.com/courts/evidence/us-dnev-2025-demissie-v-ford" }, { "decision_id": "ca-bccrt-2025-agyeman-v-hosseinzadmoghadasi", "case_name": "Agyeman v. Hosseinzadmoghadasi", "court": "British Columbia Civil Resolution Tribunal", "country": "CA", "jurisdiction": "canada", "state": "BC", "date": "2025-09-15", "citation": "2025 BCCRT 1278", "docket_number": "SC-2024-005378", "evidence_type": "document", "ai_role": "acknowledged-ai-output", "offered_by": "applicant", "challenge": "weight", "ruling": "no-weight", "rules_cited": [], "court_finding_quote": "The document was not written by an expert and is not a business record.", "text_licence": "quoted-short", "summary": "A driving-school claimant submitted an AI-generated overview of insurance policies. The CRT held the document was not expert evidence or a business record, did not prove facts, was not reliable evidence, and did not consider it when deciding the small claim.", "sensitive": false, "publication_ban_checked": false, "related_opinion_id": null, "source_url": "https://www.damiencharlotin.com/documents/792/SC-2024-005378_Decision.pdf", "archive_url": "https://web.archive.org/web/20260907221316/https://websitedc.s3.amazonaws.com/documents/SC-2024-005378_Decision.pdf", "fetched_at": "2026-09-07T00:00:00Z", "verification": "fetched-and-read", "lead_source": [ "Damien Charlotin acknowledged-AI-output lead CSV (CC0)" ], "notes": "Official CRT item URL not located; downloaded lead PDF copy was read and publication-ban status not confirmed.", "url": "https://safelegalai.com/courts/evidence/ca-bccrt-2025-agyeman-v-hosseinzadmoghadasi", "notice": "Provided as is, without warranty. Not legal advice. SafeLegalAI (Cognesio LLP) records what courts, regulators, legislatures and vendors' own public pages state; the linked official documents are the record. Names and marks belong to their owners. Anyone named may reply: https://safelegalai.com/report. Full terms: https://safelegalai.com/disclaimer", "page": "https://safelegalai.com/courts/evidence/ca-bccrt-2025-agyeman-v-hosseinzadmoghadasi" }, { "decision_id": "us-ca-alameda-2025-mendones-cushman-wakefield", "case_name": "Mendones v. Cushman & Wakefield", "court": "Superior Court of California, Alameda County", "country": "US", "jurisdiction": "us-state", "state": "CA", "date": "2025-09-09", "citation": null, "docket_number": "23CV028772", "evidence_type": "mixed", "ai_role": "ai-generated-submitted", "offered_by": "plaintiff", "challenge": "sanctions", "ruling": "sanction", "rules_cited": [ "Cal. Code Civ. Proc. § 128.7" ], "court_finding_quote": "The Court finds that exhibits 6A and 6C are products of GenAI and do not capture the actual speech and image of Geri Haas.", "text_licence": "public-domain", "summary": "Plaintiffs filed videos, photographs, and message screenshots supporting summary judgment. After an order to show cause, the court found several submissions were GenAI deepfakes or materially altered, rejected unreliable metadata and explanations, struck the complaint, dismissed with prejudice, and declined criminal referral.", "sensitive": false, "publication_ban_checked": false, "related_opinion_id": "cal-super-2025-mendones-v-cushman-wakefield", "source_url": "https://www.damiencharlotin.com/documents/2960/mendones-v-cushman-wakefield_EPA7JBJ.pdf", "archive_url": null, "fetched_at": "2026-09-07T00:00:00Z", "verification": "link-only", "lead_source": [ "Damien Charlotin deepfakes lead CSV (CC0)", "User brief named-case list" ], "notes": "No court-hosted copy was located; Reason/CDN copy and lead PDF were read. The order includes media links that are intentionally not reused.", "url": "https://safelegalai.com/courts/evidence/us-ca-alameda-2025-mendones-cushman-wakefield", "notice": "Provided as is, without warranty. Not legal advice. SafeLegalAI (Cognesio LLP) records what courts, regulators, legislatures and vendors' own public pages state; the linked official documents are the record. Names and marks belong to their owners. Anyone named may reply: https://safelegalai.com/report. Full terms: https://safelegalai.com/disclaimer", "page": "https://safelegalai.com/courts/evidence/us-ca-alameda-2025-mendones-cushman-wakefield" }, { "decision_id": "uk-ukftt-2025-vp-evans-v-hmrc", "case_name": "VP Evans Ors v. HMRC", "court": "First-tier Tribunal (Tax Chamber)", "country": "GB", "jurisdiction": "uk", "state": null, "date": "2025-08-22", "citation": "[2025] UKFTT 1112 (TC)", "docket_number": "TC09638", "evidence_type": "document", "ai_role": "acknowledged-ai-output", "offered_by": "court", "challenge": "none", "ruling": "other", "rules_cited": [ "Tribunal Procedure Rules" ], "court_finding_quote": "Any use of AI by or on behalf of the judiciary must protect the integrity of the administration of justice.", "text_licence": "quoted-short", "summary": "The tribunal disclosed remarks on judicial AI use in a case-management decision, referencing official judiciary guidance and eJudiciary Copilot Chat. The AI was discussed as a secure summarisation or support tool for judicial office holders, not as party evidence determining contested facts.", "sensitive": false, "publication_ban_checked": true, "related_opinion_id": null, "source_url": "https://caselaw.nationalarchives.gov.uk/ukftt/tc/2025/1112", "archive_url": "https://web.archive.org/web/20260907221332/https://caselaw.nationalarchives.gov.uk/ukftt/tc/2025/1112", "fetched_at": "2026-09-07T00:00:00Z", "verification": "fetched-and-read", "lead_source": [ "Damien Charlotin acknowledged-AI-output lead CSV (CC0)" ], "notes": null, "url": "https://safelegalai.com/courts/evidence/uk-ukftt-2025-vp-evans-v-hmrc", "notice": "Provided as is, without warranty. Not legal advice. SafeLegalAI (Cognesio LLP) records what courts, regulators, legislatures and vendors' own public pages state; the linked official documents are the record. Names and marks belong to their owners. Anyone named may reply: https://safelegalai.com/report. Full terms: https://safelegalai.com/disclaimer", "page": "https://safelegalai.com/courts/evidence/uk-ukftt-2025-vp-evans-v-hmrc" }, { "decision_id": "nz-ca-2025-kipping-v-ewrb", "case_name": "Kipping v. Electrical Workers Registration Board", "court": "Court of Appeal of New Zealand", "country": "NZ", "jurisdiction": "nz", "state": null, "date": "2025-08-20", "citation": null, "docket_number": null, "evidence_type": "chat-output", "ai_role": "acknowledged-ai-output", "offered_by": "applicant", "challenge": "weight", "ruling": "no-weight", "rules_cited": [], "court_finding_quote": "A submission relaying an answer given by artificial intelligence, without more, is not reliable evidence.", "text_licence": "quoted-short", "summary": "A self-represented appellant cited a Microsoft Copilot answer about isolator switches. The Court of Appeal held the AI answer took him no further because an unattributed chatbot response, without supporting evidence, was not reliable proof on the technical issue.", "sensitive": false, "publication_ban_checked": true, "related_opinion_id": null, "source_url": "https://www.justice.govt.nz/jdo_documents/workspace___SpacesStore_0438ea8f_2b79_407a_9b2b_bd23663dcb54.pdf", "archive_url": "http://web.archive.org/web/20260509200414/https://www.justice.govt.nz/jdo_documents/workspace___SpacesStore_0438ea8f_2b79_407a_9b2b_bd23663dcb54.pdf", "fetched_at": "2026-09-07T00:00:00Z", "verification": "fetched-and-read", "lead_source": [ "Damien Charlotin acknowledged-AI-output lead CSV (CC0)" ], "notes": null, "url": "https://safelegalai.com/courts/evidence/nz-ca-2025-kipping-v-ewrb", "notice": "Provided as is, without warranty. Not legal advice. SafeLegalAI (Cognesio LLP) records what courts, regulators, legislatures and vendors' own public pages state; the linked official documents are the record. Names and marks belong to their owners. Anyone named may reply: https://safelegalai.com/report. Full terms: https://safelegalai.com/disclaimer", "page": "https://safelegalai.com/courts/evidence/nz-ca-2025-kipping-v-ewrb" }, { "decision_id": "ca-bccrt-2025-brookbrand-v-hemp-hemp-horray", "case_name": "Brookbrand Wholesale Ltd. v. Hemp Hemp Horray Holdings Ltd", "court": "British Columbia Civil Resolution Tribunal", "country": "CA", "jurisdiction": "canada", "state": "BC", "date": "2025-07-25", "citation": null, "docket_number": null, "evidence_type": "chat-output", "ai_role": "acknowledged-ai-output", "offered_by": "defendant", "challenge": "weight", "ruling": "no-weight", "rules_cited": [], "court_finding_quote": "While a useful research tool, AI is also unreliable and sometimes provides inaccurate information.", "text_licence": "quoted-short", "summary": "Hemp Hemp Horray relied on screenshots of AI summaries about CBD oil. The CRT recognized AI can be a research tool but found unverified AI summaries unreliable and gave them little weight, deciding the dispute based on admissible evidence instead.", "sensitive": false, "publication_ban_checked": false, "related_opinion_id": null, "source_url": "https://decisions.civilresolutionbc.ca/crt/en/nav.do", "archive_url": "https://web.archive.org/web/20260907221411/https://decisions.civilresolutionbc.ca/crt/en/nav.do", "fetched_at": "2026-09-07T00:00:00Z", "verification": "link-only", "lead_source": [ "Damien Charlotin acknowledged-AI-output lead CSV (CC0)" ], "notes": "No exact CRT item URL located; row coded from lead excerpt. Publication-ban status not confirmed.", "url": "https://safelegalai.com/courts/evidence/ca-bccrt-2025-brookbrand-v-hemp-hemp-horray", "notice": "Provided as is, without warranty. Not legal advice. SafeLegalAI (Cognesio LLP) records what courts, regulators, legislatures and vendors' own public pages state; the linked official documents are the record. Names and marks belong to their owners. Anyone named may reply: https://safelegalai.com/report. Full terms: https://safelegalai.com/disclaimer", "page": "https://safelegalai.com/courts/evidence/ca-bccrt-2025-brookbrand-v-hemp-hemp-horray" }, { "decision_id": "it-brescia-2025-p-v-cp", "case_name": "P v. CP", "court": "Tribunale di Brescia", "country": "IT", "jurisdiction": "other", "state": null, "date": "2025-07-22", "citation": null, "docket_number": null, "evidence_type": "chat-output", "ai_role": "acknowledged-ai-output", "offered_by": "defendant", "challenge": "weight", "ruling": "no-weight", "rules_cited": [ "Italian civil procedure precautionary measures" ], "court_finding_quote": "Le conclusioni di ChatGPT, prive di riferimenti a dati specifici, non sono verificabili.", "text_licence": "quoted-short", "summary": "A company attempted to resist conservatory seizure with a ChatGPT-generated financial analysis purporting to show solvency. The Brescia court found the output lacked specific verifiable data, gave it no meaningful evidentiary value, and upheld the precautionary measure.", "sensitive": false, "publication_ban_checked": false, "related_opinion_id": null, "source_url": "https://www.damiencharlotin.com/documents/979/brescia-chatgpt.pdf", "archive_url": "https://web.archive.org/web/20260907221456/https://websitedc.s3.amazonaws.com/documents/brescia-chatgpt.pdf", "fetched_at": "2026-09-07T00:00:00Z", "verification": "fetched-and-read", "lead_source": [ "Damien Charlotin acknowledged-AI-output lead CSV (CC0)" ], "notes": "No official court URL located; publication-ban status not confirmed.", "url": "https://safelegalai.com/courts/evidence/it-brescia-2025-p-v-cp", "notice": "Provided as is, without warranty. Not legal advice. SafeLegalAI (Cognesio LLP) records what courts, regulators, legislatures and vendors' own public pages state; the linked official documents are the record. Names and marks belong to their owners. Anyone named may reply: https://safelegalai.com/report. Full terms: https://safelegalai.com/disclaimer", "page": "https://safelegalai.com/courts/evidence/it-brescia-2025-p-v-cp" }, { "decision_id": "au-qcat-2025-marshall-v-regal-waters", "case_name": "Marshall v. Regal Waters Holdings Pty Ltd", "court": "Queensland Civil and Administrative Tribunal", "country": "AU", "jurisdiction": "australia", "state": "QLD", "date": "2025-07-22", "citation": "[2025] QCAT 289", "docket_number": null, "evidence_type": "chat-output", "ai_role": "acknowledged-ai-output", "offered_by": "applicant", "challenge": "weight", "ruling": "no-weight", "rules_cited": [], "court_finding_quote": "Unverified, unattributed and untested statements generated by Google’s AI applications have no probative value.", "text_licence": "quoted-short", "summary": "Homeowners cited a Google AI response about rent assistance and housing stress in a community-title dispute. QCAT held the generated statements were unverified, unattributed, and untested, assigned no probative value, and rejected the alternative calculation for lack of credible evidence.", "sensitive": false, "publication_ban_checked": false, "related_opinion_id": null, "source_url": "https://www.austlii.edu.au/cgi-bin/viewdoc/au/cases/qld/QCAT/2025/289.html", "archive_url": "http://web.archive.org/web/20251002175137/https://www.austlii.edu.au/cgi-bin/viewdoc/au/cases/qld/QCAT/2025/289.html", "fetched_at": "2026-09-07T00:00:00Z", "verification": "fetched-and-read", "lead_source": [ "Damien Charlotin acknowledged-AI-output lead CSV (CC0)" ], "notes": "AustLII live page blocked automated reading; quote and coding came from lead excerpt. Publication-ban status not confirmed.", "url": "https://safelegalai.com/courts/evidence/au-qcat-2025-marshall-v-regal-waters", "notice": "Provided as is, without warranty. Not legal advice. SafeLegalAI (Cognesio LLP) records what courts, regulators, legislatures and vendors' own public pages state; the linked official documents are the record. Names and marks belong to their owners. Anyone named may reply: https://safelegalai.com/report. Full terms: https://safelegalai.com/disclaimer", "page": "https://safelegalai.com/courts/evidence/au-qcat-2025-marshall-v-regal-waters" }, { "decision_id": "us-edmi-2025-hunt-v-morissette", "case_name": "Hunt v. Morissette", "court": "U.S. District Court for the Eastern District of Michigan", "country": "US", "jurisdiction": "us-federal", "state": "MI", "date": "2025-06-11", "citation": null, "docket_number": "2:24-cv-12947", "evidence_type": "chat-output", "ai_role": "acknowledged-ai-output", "offered_by": "plaintiff", "challenge": "weight", "ruling": "no-weight", "rules_cited": [ "Fed. R. Civ. P. 12", "Fed. R. Civ. P. 11" ], "court_finding_quote": "Because AI-generated material is unlikely to be helpful to the Court, Hunt should not include or rely upon AI-generated material.", "text_licence": "public-domain", "summary": "A pro se plaintiff relied on Ask.AI-style legal material in opposing dismissal. The court explained that AI outputs can be unreliable and unhelpful, rejected reliance on them, dismissed claims, and warned that future pleadings should not include or rely on AI-generated legal argument.", "sensitive": false, "publication_ban_checked": true, "related_opinion_id": "edmi-2025-hunt-v-morissette", "source_url": "https://www.damiencharlotin.com/documents/577/Hunt_v._Morissette_et_al_USA_6_June_2025.pdf", "archive_url": "https://web.archive.org/web/20260817232955/https://websitedc.s3.amazonaws.com/documents/Hunt_v._Morissette_et_al_USA_6_June_2025.pdf", "fetched_at": "2026-09-07T00:00:00Z", "verification": "fetched-and-read", "lead_source": [ "Damien Charlotin acknowledged-AI-output lead CSV (CC0)" ], "notes": null, "url": "https://safelegalai.com/courts/evidence/us-edmi-2025-hunt-v-morissette", "notice": "Provided as is, without warranty. Not legal advice. SafeLegalAI (Cognesio LLP) records what courts, regulators, legislatures and vendors' own public pages state; the linked official documents are the record. Names and marks belong to their owners. Anyone named may reply: https://safelegalai.com/report. Full terms: https://safelegalai.com/disclaimer", "page": "https://safelegalai.com/courts/evidence/us-edmi-2025-hunt-v-morissette" }, { "decision_id": "ca-oncj-2025-r-v-cheng", "case_name": "R. v. Cheng", "court": "Ontario Court of Justice", "country": "CA", "jurisdiction": "canada", "state": "ON", "date": "2025-05-07", "citation": "2025 ONCJ 252", "docket_number": null, "evidence_type": "mixed", "ai_role": "alleged-deepfake", "offered_by": "prosecution", "challenge": "authenticity", "ruling": "admitted", "rules_cited": [ "Canada Evidence Act s.31.1" ], "court_finding_quote": "Credible witnesses and system evidence explained the gap; the deepfake concern was unsupported.", "text_licence": "quoted-short", "summary": "Cheng challenged a home-surveillance video and screenshot as possibly altered or misdated. The court accepted testimony from the former partner and daughter that the files came from the system’s SIM card and were uploaded unchanged, found the gap explainable, admitted the exhibits, and convicted him.", "sensitive": false, "publication_ban_checked": false, "related_opinion_id": null, "source_url": "https://www.canlii.org/en/on/oncj/doc/2025/2025oncj252/2025oncj252.html", "archive_url": null, "fetched_at": "2026-09-07T00:00:00Z", "verification": "fetched-and-read", "lead_source": [ "Damien Charlotin deepfakes lead CSV (CC0)", "User brief named-case list" ], "notes": "CanLII page could not be independently read by script; downloaded lead PDF copy was read.", "url": "https://safelegalai.com/courts/evidence/ca-oncj-2025-r-v-cheng", "notice": "Provided as is, without warranty. Not legal advice. SafeLegalAI (Cognesio LLP) records what courts, regulators, legislatures and vendors' own public pages state; the linked official documents are the record. Names and marks belong to their owners. Anyone named may reply: https://safelegalai.com/report. Full terms: https://safelegalai.com/disclaimer", "page": "https://safelegalai.com/courts/evidence/ca-oncj-2025-r-v-cheng" }, { "decision_id": "us-az-super-2025-state-v-horcasitas", "case_name": "State v. Horcasitas", "court": "Superior Court of Arizona, Maricopa County", "country": "US", "jurisdiction": "us-state", "state": "AZ", "date": "2025-05-01", "citation": null, "docket_number": null, "evidence_type": "victim-statement", "ai_role": "ai-generated-submitted", "offered_by": "prosecution", "challenge": "admissibility", "ruling": "admitted", "rules_cited": [ "Arizona victims' rights law", "Due process" ], "court_finding_quote": "The sentencing court allowed an AI-generated avatar victim-impact statement before imposing a 10.5-year sentence.", "text_licence": "public-domain", "summary": "At sentencing for a road-rage killing, the court allowed the victim’s family to present an AI-generated avatar delivering a scripted victim-impact statement. The defense noticed an appeal; secondary searches found no Arizona appellate ruling by September 2026, so notes record the posture as pending.", "sensitive": true, "publication_ban_checked": false, "related_opinion_id": "az-super-2025-state-v-horcasitas", "source_url": "https://www.abc15.com/news/region-southeast-valley/chandler/defense-attorney-appeals-after-ai-video-used-in-court-sentencing", "archive_url": "http://web.archive.org/web/20260907220942/https://www.abc15.com/news/region-southeast-valley/chandler/defense-attorney-appeals-after-ai-video-used-in-court-sentencing", "fetched_at": "2026-09-07T00:00:00Z", "verification": "link-only", "lead_source": [ "User brief named-case list" ], "notes": "No official sentencing transcript or appellate ruling located. Secondary reports state appeal was filed and remained unresolved in public reporting as of September 2026; publication-ban status not confirmed.", "url": "https://safelegalai.com/courts/evidence/us-az-super-2025-state-v-horcasitas", "notice": "Provided as is, without warranty. Not legal advice. SafeLegalAI (Cognesio LLP) records what courts, regulators, legislatures and vendors' own public pages state; the linked official documents are the record. Names and marks belong to their owners. Anyone named may reply: https://safelegalai.com/report. Full terms: https://safelegalai.com/disclaimer", "page": "https://safelegalai.com/courts/evidence/us-az-super-2025-state-v-horcasitas" }, { "decision_id": "au-vcat-2025-metcalf-v-hector", "case_name": "Metcalf v Hector", "court": "Victorian Civil and Administrative Tribunal", "country": "AU", "jurisdiction": "australia", "state": "VIC", "date": "2025-04-29", "citation": "[2025] VCAT 367", "docket_number": null, "evidence_type": "chat-output", "ai_role": "acknowledged-ai-output", "offered_by": "respondent", "challenge": "weight", "ruling": "no-weight", "rules_cited": [], "court_finding_quote": "The Gemini search about standard fence heights was not of assistance in determining the claims.", "text_licence": "quoted-short", "summary": "A respondent submitted a Gemini search result about standard fence heights in Australia. VCAT declined to rely on the chatbot statement, finding it did not assist in resolving the factual dispute, and decided the neighbourhood fencing claim on conventional evidence.", "sensitive": false, "publication_ban_checked": false, "related_opinion_id": null, "source_url": "https://www.austlii.edu.au/cgi-bin/viewdoc/au/cases/vic/VCAT/2025/367.html", "archive_url": "http://web.archive.org/web/20260307013844/https://www.austlii.edu.au/cgi-bin/viewdoc/au/cases/vic/VCAT/2025/367.html", "fetched_at": "2026-09-07T00:00:00Z", "verification": "fetched-and-read", "lead_source": [ "Damien Charlotin acknowledged-AI-output lead CSV (CC0)" ], "notes": "AustLII live page blocked automated reading; quote and coding came from lead excerpt. Publication-ban status not confirmed.", "url": "https://safelegalai.com/courts/evidence/au-vcat-2025-metcalf-v-hector", "notice": "Provided as is, without warranty. Not legal advice. SafeLegalAI (Cognesio LLP) records what courts, regulators, legislatures and vendors' own public pages state; the linked official documents are the record. Names and marks belong to their owners. Anyone named may reply: https://safelegalai.com/report. Full terms: https://safelegalai.com/disclaimer", "page": "https://safelegalai.com/courts/evidence/au-vcat-2025-metcalf-v-hector" }, { "decision_id": "us-edny-2025-ferlito-v-harbor-freight", "case_name": "Ferlito v. Harbor Freight Tools USA", "court": "U.S. District Court for the Eastern District of New York", "country": "US", "jurisdiction": "us-federal", "state": "NY", "date": "2025-04-23", "citation": null, "docket_number": null, "evidence_type": "expert-report", "ai_role": "acknowledged-ai-output", "offered_by": "plaintiff", "challenge": "admissibility", "ruling": "admitted", "rules_cited": [ "Fed. R. Evid. 702", "Daubert" ], "court_finding_quote": "Lehnert's post hoc use of ChatGPT did not render his opinions unreliable because he had reached his conclusions independently.", "text_licence": "public-domain", "summary": "Plaintiff’s safety expert used ChatGPT only after forming his opinions to see whether it agreed with his alternative-design theory. The court distinguished post hoc checking from reliance, found his practical experience and methodology sufficient, and denied Harbor Freight’s Daubert exclusion motion.", "sensitive": false, "publication_ban_checked": true, "related_opinion_id": "edny-2025-ferlito-v-harbor-freight", "source_url": "https://www.damiencharlotin.com/documents/350/Ferlito_v._Harbor_Freight_Tools_USA_Inc._E.D.N.Y._April_23_2025.pdf", "archive_url": null, "fetched_at": "2026-09-07T00:00:00Z", "verification": "fetched-and-read", "lead_source": [ "Damien Charlotin acknowledged-AI-output lead CSV (CC0)" ], "notes": null, "url": "https://safelegalai.com/courts/evidence/us-edny-2025-ferlito-v-harbor-freight", "notice": "Provided as is, without warranty. Not legal advice. SafeLegalAI (Cognesio LLP) records what courts, regulators, legislatures and vendors' own public pages state; the linked official documents are the record. Names and marks belong to their owners. Anyone named may reply: https://safelegalai.com/report. Full terms: https://safelegalai.com/disclaimer", "page": "https://safelegalai.com/courts/evidence/us-edny-2025-ferlito-v-harbor-freight" }, { "decision_id": "ca-oncj-2025-r-v-vitellaro", "case_name": "R. v. Vitellaro", "court": "Ontario Court of Justice", "country": "CA", "jurisdiction": "canada", "state": "ON", "date": "2025-04-11", "citation": "2025 ONCJ 200", "docket_number": null, "evidence_type": "video", "ai_role": "alleged-deepfake", "offered_by": "prosecution", "challenge": "authenticity", "ruling": "admitted", "rules_cited": [ "Canada Evidence Act s.31.1" ], "court_finding_quote": "Three USB copies were compared and found consistent; the CGI deepfake allegation did not displace admission.", "text_licence": "quoted-short", "summary": "Vitellaro alleged a police booking-cell video was altered and called it a CGI deepfake. In court, the parties compared three USB copies, including the alleged original, and the judge inspected the footage, found the copies consistent, and rejected the authenticity challenge.", "sensitive": false, "publication_ban_checked": false, "related_opinion_id": null, "source_url": "https://www.canlii.org/en/on/oncj/doc/2025/2025oncj200/2025oncj200.html", "archive_url": null, "fetched_at": "2026-09-07T00:00:00Z", "verification": "fetched-and-read", "lead_source": [ "Damien Charlotin deepfakes lead CSV (CC0)", "User brief named-case list" ], "notes": "CanLII page could not be independently read by script; downloaded lead PDF copy was read.", "url": "https://safelegalai.com/courts/evidence/ca-oncj-2025-r-v-vitellaro", "notice": "Provided as is, without warranty. Not legal advice. SafeLegalAI (Cognesio LLP) records what courts, regulators, legislatures and vendors' own public pages state; the linked official documents are the record. Names and marks belong to their owners. Anyone named may reply: https://safelegalai.com/report. Full terms: https://safelegalai.com/disclaimer", "page": "https://safelegalai.com/courts/evidence/ca-oncj-2025-r-v-vitellaro" }, { "decision_id": "ca-bccrt-2025-bahia-v-icbc", "case_name": "Bahia v. ICBC", "court": "British Columbia Civil Resolution Tribunal", "country": "CA", "jurisdiction": "canada", "state": "BC", "date": "2025-04-10", "citation": "2025 BCCRT 453", "docket_number": "AB-2023-012674", "evidence_type": "chat-output", "ai_role": "acknowledged-ai-output", "offered_by": "applicant", "challenge": "weight", "ruling": "no-weight", "rules_cited": [], "court_finding_quote": "I give no weight to Mr. Bahia’s AI-produced evidence about the risks.", "text_licence": "quoted-short", "summary": "An accident-benefits applicant submitted Google AI and other online material about health risks. The CRT followed prior skepticism of unregulated AI outputs, found the material unreliable at best, and gave no weight to the AI-produced evidence.", "sensitive": true, "publication_ban_checked": false, "related_opinion_id": null, "source_url": "https://www.damiencharlotin.com/documents/574/Bahia_v._ICBC_Canada_10_April_2025.pdf", "archive_url": null, "fetched_at": "2026-09-07T00:00:00Z", "verification": "fetched-and-read", "lead_source": [ "Damien Charlotin acknowledged-AI-output lead CSV (CC0)" ], "notes": "Official CRT item URL not located; downloaded lead PDF copy was read and publication-ban status not confirmed.", "url": "https://safelegalai.com/courts/evidence/ca-bccrt-2025-bahia-v-icbc", "notice": "Provided as is, without warranty. Not legal advice. SafeLegalAI (Cognesio LLP) records what courts, regulators, legislatures and vendors' own public pages state; the linked official documents are the record. Names and marks belong to their owners. Anyone named may reply: https://safelegalai.com/report. Full terms: https://safelegalai.com/disclaimer", "page": "https://safelegalai.com/courts/evidence/ca-bccrt-2025-bahia-v-icbc" }, { "decision_id": "us-dcca-2025-ross-v-united-states", "case_name": "Ross v. United States", "court": "District of Columbia Court of Appeals", "country": "US", "jurisdiction": "us-state", "state": "DC", "date": "2025-02-20", "citation": null, "docket_number": null, "evidence_type": "chat-output", "ai_role": "acknowledged-ai-output", "offered_by": "court", "challenge": "none", "ruling": "other", "rules_cited": [ "D.C. Code § 22-1001" ], "court_finding_quote": "The dissent used ChatGPT to test common-sense inferences, but the majority reversed for insufficient evidence.", "text_licence": "public-domain", "summary": "A dissenting judge used ChatGPT queries to compare intuitive inferences about animal welfare, while the majority held the government’s evidence insufficient without proof of vehicle temperature or distress. The AI-aided reasoning did not affect the judgment, which reversed the conviction.", "sensitive": false, "publication_ban_checked": true, "related_opinion_id": "dccoa-2025-ross-v-united-states", "source_url": "https://reason.com/volokh/2025/03/25/d-c-judge-uses-chatgpt-in-discussing-whether-common-knowledge-inference-in-criminal-case-was-justified/", "archive_url": "https://web.archive.org/web/20260907221607/https://reason.com/volokh/2025/03/25/d-c-judge-uses-chatgpt-in-discussing-whether-common-knowledge-inference-in-criminal-case-was-justified/", "fetched_at": "2026-09-07T00:00:00Z", "verification": "link-only", "lead_source": [ "Damien Charlotin acknowledged-AI-output lead CSV (CC0)" ], "notes": "Official D.C. Court of Appeals opinion URL not located in this session; coded from legal commentary and lead text.", "url": "https://safelegalai.com/courts/evidence/us-dcca-2025-ross-v-united-states", "notice": "Provided as is, without warranty. Not legal advice. SafeLegalAI (Cognesio LLP) records what courts, regulators, legislatures and vendors' own public pages state; the linked official documents are the record. Names and marks belong to their owners. Anyone named may reply: https://safelegalai.com/report. Full terms: https://safelegalai.com/disclaimer", "page": "https://safelegalai.com/courts/evidence/us-dcca-2025-ross-v-united-states" }, { "decision_id": "nl-rvs-2025-aleto-beheer-v-venlo", "case_name": "Aleto Beheer BV v. Venlo Municipality", "court": "Council of State of the Netherlands", "country": "NL", "jurisdiction": "other", "state": null, "date": "2025-01-29", "citation": "ECLI:NL:RVS:2025:335", "docket_number": null, "evidence_type": "chat-output", "ai_role": "acknowledged-ai-output", "offered_by": "applicant", "challenge": "weight", "ruling": "no-weight", "rules_cited": [], "court_finding_quote": "ChatGPT statements without the prompt, sources, or verification could not replace an independent valuation expert.", "text_licence": "quoted-short", "summary": "Aleto filed a late supplementary document with ChatGPT-based claims about environmental zoning and real-estate values. The Dutch Council of State refused to treat the opaque, unverified output as evidence in a technical valuation dispute and dismissed the appeal.", "sensitive": false, "publication_ban_checked": false, "related_opinion_id": null, "source_url": "https://uitspraken.rechtspraak.nl/details?id=ECLI:NL:RVS:2025:335", "archive_url": "https://web.archive.org/web/20260907221636/https://uitspraken.rechtspraak.nl/details?id=ECLI:NL:RVS:2025:335", "fetched_at": "2026-09-07T00:00:00Z", "verification": "fetched-and-read", "lead_source": [ "Damien Charlotin acknowledged-AI-output lead CSV (CC0)" ], "notes": "Official Rechtspraak URL was identified from the PDF but failed TLS fetch; downloaded lead PDF copy was read. Publication-ban status not confirmed.", "url": "https://safelegalai.com/courts/evidence/nl-rvs-2025-aleto-beheer-v-venlo", "notice": "Provided as is, without warranty. Not legal advice. SafeLegalAI (Cognesio LLP) records what courts, regulators, legislatures and vendors' own public pages state; the linked official documents are the record. Names and marks belong to their owners. Anyone named may reply: https://safelegalai.com/report. Full terms: https://safelegalai.com/disclaimer", "page": "https://safelegalai.com/courts/evidence/nl-rvs-2025-aleto-beheer-v-venlo" }, { "decision_id": "il-mag-2024-lhs-v-clal-insurance", "case_name": "L.H.S. v. Clal Insurance Co.", "court": "Magistrate Court of Israel", "country": "IL", "jurisdiction": "israel", "state": null, "date": "2024-12-09", "citation": null, "docket_number": null, "evidence_type": "document", "ai_role": "ai-generated-submitted", "offered_by": "defendant", "challenge": "admissibility", "ruling": "excluded", "rules_cited": [], "court_finding_quote": "The expert was ordered to disregard the AI-generated summary and marked medical documents.", "text_licence": "quoted-short", "summary": "An insurer sent a court-appointed medical expert a DigitalOwl AI summary and highlighted medical records. The court treated the processed content as risking influence on the neutral expert, ordered clean unmarked records resent, and denied costs because the issue was novel.", "sensitive": true, "publication_ban_checked": false, "related_opinion_id": null, "source_url": "https://www.damiencharlotin.com/documents/462/L._H._S._v._Clal_Insurance_Co._Israel_9_December_2024.pdf", "archive_url": null, "fetched_at": "2026-09-07T00:00:00Z", "verification": "fetched-and-read", "lead_source": [ "Damien Charlotin acknowledged-AI-output lead CSV (CC0)", "User brief named-case list" ], "notes": "Hebrew PDF copy read; no official source or publication-ban status confirmed.", "url": "https://safelegalai.com/courts/evidence/il-mag-2024-lhs-v-clal-insurance", "notice": "Provided as is, without warranty. Not legal advice. SafeLegalAI (Cognesio LLP) records what courts, regulators, legislatures and vendors' own public pages state; the linked official documents are the record. Names and marks belong to their owners. Anyone named may reply: https://safelegalai.com/report. Full terms: https://safelegalai.com/disclaimer", "page": "https://safelegalai.com/courts/evidence/il-mag-2024-lhs-v-clal-insurance" }, { "decision_id": "us-pa-super-2024-in-re-ds", "case_name": "In re D.S.", "court": "Superior Court of Pennsylvania", "country": "US", "jurisdiction": "us-state", "state": "PA", "date": "2024-11-14", "citation": null, "docket_number": "2032 EDA 2023", "evidence_type": "video", "ai_role": "alleged-deepfake", "offered_by": "prosecution", "challenge": "authenticity", "ruling": "admitted", "rules_cited": [ "Pa.R.A.P. 302(a)", "Pa.R.E. 901" ], "court_finding_quote": "Because D.S. did not object at trial on deepfake or manipulation grounds, the appellate court held the claim waived.", "text_licence": "public-domain", "summary": "The Commonwealth used an Instagram Live video allegedly showing D.S. holding a firearm. On appeal he suggested manipulation or deepfake concerns, but the Superior Court treated those arguments as waived because they were not raised at trial; authentication rested on witness recognition and platform explanation.", "sensitive": true, "publication_ban_checked": true, "related_opinion_id": "pa-super-2024-in-re-ds", "source_url": "https://www.damiencharlotin.com/documents/3003/2024-2032-eda-2023.pdf", "archive_url": null, "fetched_at": "2026-09-07T00:00:00Z", "verification": "fetched-and-read", "lead_source": [ "Damien Charlotin deepfakes lead CSV (CC0)", "User brief named-case list" ], "notes": null, "url": "https://safelegalai.com/courts/evidence/us-pa-super-2024-in-re-ds", "notice": "Provided as is, without warranty. Not legal advice. SafeLegalAI (Cognesio LLP) records what courts, regulators, legislatures and vendors' own public pages state; the linked official documents are the record. Names and marks belong to their owners. Anyone named may reply: https://safelegalai.com/report. Full terms: https://safelegalai.com/disclaimer", "page": "https://safelegalai.com/courts/evidence/us-pa-super-2024-in-re-ds" }, { "decision_id": "be-rvsce-2024-wallonie-bruxelles", "case_name": "XXXX v. Wallonie Bruxelles", "court": "Council of State of Belgium", "country": "BE", "jurisdiction": "other", "state": null, "date": "2024-11-08", "citation": "ECLI:BE:RVSCE:2024:ARR.261.320", "docket_number": null, "evidence_type": "chat-output", "ai_role": "acknowledged-ai-output", "offered_by": "applicant", "challenge": "weight", "ruling": "no-weight", "rules_cited": [], "court_finding_quote": "Le délai de traitement renseigné par ChatGPT n'établissait pas l'extrême urgence requise.", "text_licence": "quoted-short", "summary": "Counsel cited ChatGPT estimates of processing time to support an emergency suspension request. The Belgian Council of State did not accept that generated timing information as proof of urgency and rejected the request on traditional irreparable-harm grounds.", "sensitive": false, "publication_ban_checked": false, "related_opinion_id": null, "source_url": "https://www.damiencharlotin.com/documents/887/ECLI_BE_RVSCE_2024_ARR.261.320_FR.pdf", "archive_url": "https://web.archive.org/web/20260817232955/https://websitedc.s3.amazonaws.com/documents/ECLI_BE_RVSCE_2024_ARR.261.320_FR.pdf", "fetched_at": "2026-09-07T00:00:00Z", "verification": "fetched-and-read", "lead_source": [ "Damien Charlotin acknowledged-AI-output lead CSV (CC0)" ], "notes": "French PDF copy read; no official source or publication-ban status confirmed.", "url": "https://safelegalai.com/courts/evidence/be-rvsce-2024-wallonie-bruxelles", "notice": "Provided as is, without warranty. Not legal advice. SafeLegalAI (Cognesio LLP) records what courts, regulators, legislatures and vendors' own public pages state; the linked official documents are the record. Names and marks belong to their owners. Anyone named may reply: https://safelegalai.com/report. Full terms: https://safelegalai.com/disclaimer", "page": "https://safelegalai.com/courts/evidence/be-rvsce-2024-wallonie-bruxelles" }, { "decision_id": "nl-rbdha-2024-plaintiff-v-minister-asylum", "case_name": "Plaintiff v. Minister of Asylum", "court": "District Court of The Hague", "country": "NL", "jurisdiction": "other", "state": null, "date": "2024-11-06", "citation": "ECLI:NL:RBDHA:2024:18167", "docket_number": null, "evidence_type": "chat-output", "ai_role": "acknowledged-ai-output", "offered_by": "applicant", "challenge": "weight", "ruling": "no-weight", "rules_cited": [], "court_finding_quote": "Unverifiable ChatGPT material could not substantiate claims about Moroccan surveillance practices.", "text_licence": "quoted-short", "summary": "An asylum representative invoked ChatGPT output to claim Moroccan authorities monitor dissidents abroad. The court discounted the unsupported chatbot assertion because the prompt, answer, sources, and independent corroboration were not provided, and dismissed the asylum appeal.", "sensitive": true, "publication_ban_checked": false, "related_opinion_id": null, "source_url": "https://uitspraken.rechtspraak.nl/details?id=ECLI:NL:RBDHA:2024:18167", "archive_url": null, "fetched_at": "2026-09-07T00:00:00Z", "verification": "fetched-and-read", "lead_source": [ "Damien Charlotin acknowledged-AI-output lead CSV (CC0)" ], "notes": "Official Rechtspraak URL was identified from the PDF but failed TLS fetch; downloaded lead PDF copy was read. Publication-ban status not confirmed.", "url": "https://safelegalai.com/courts/evidence/nl-rbdha-2024-plaintiff-v-minister-asylum", "notice": "Provided as is, without warranty. Not legal advice. SafeLegalAI (Cognesio LLP) records what courts, regulators, legislatures and vendors' own public pages state; the linked official documents are the record. Names and marks belong to their owners. Anyone named may reply: https://safelegalai.com/report. Full terms: https://safelegalai.com/disclaimer", "page": "https://safelegalai.com/courts/evidence/nl-rbdha-2024-plaintiff-v-minister-asylum" }, { "decision_id": "ca-fpslreb-2024-paynter-cbsa", "case_name": "Paynter v. Canada Border Services Agency", "court": "Federal Public Sector Labour Relations and Employment Board", "country": "CA", "jurisdiction": "canada", "state": null, "date": "2024-08-19", "citation": "2024 FPSLREB 116", "docket_number": null, "evidence_type": "video", "ai_role": "alleged-deepfake", "offered_by": "respondent", "challenge": "authenticity", "ruling": "admitted", "rules_cited": [ "Canada Labour Code", "Canada Evidence Act s.31.1" ], "court_finding_quote": "There was no evidence the CCTV footage was manipulated; contemporaneous reports and witnesses supported the timeline.", "text_licence": "quoted-short", "summary": "Paynter contended CCTV footage of a workplace entrance altercation was a deepfake. The Board reviewed the video alongside security incident reports, an investigator’s report, timestamps, and witness testimony, found the allegation unsupported, relied on the footage, and denied the grievances challenging discipline and termination.", "sensitive": false, "publication_ban_checked": false, "related_opinion_id": null, "source_url": "https://www.canlii.org/en/ca/pslreb/doc/2024/2024fpslreb116/2024fpslreb116.html", "archive_url": null, "fetched_at": "2026-09-07T00:00:00Z", "verification": "fetched-and-read", "lead_source": [ "Damien Charlotin deepfakes lead CSV (CC0)", "User brief named-case list" ], "notes": "CanLII-equivalent source could not be independently read by script; downloaded lead PDF copy was read.", "url": "https://safelegalai.com/courts/evidence/ca-fpslreb-2024-paynter-cbsa", "notice": "Provided as is, without warranty. Not legal advice. SafeLegalAI (Cognesio LLP) records what courts, regulators, legislatures and vendors' own public pages state; the linked official documents are the record. Names and marks belong to their owners. Anyone named may reply: https://safelegalai.com/report. Full terms: https://safelegalai.com/disclaimer", "page": "https://safelegalai.com/courts/evidence/ca-fpslreb-2024-paynter-cbsa" }, { "decision_id": "nl-rbgel-2024-x-v-y", "case_name": "X v. Y", "court": "District Court of Gelderland", "country": "NL", "jurisdiction": "other", "state": null, "date": "2024-07-26", "citation": "ECLI:NL:RBGEL:2024:4474", "docket_number": null, "evidence_type": "chat-output", "ai_role": "acknowledged-ai-output", "offered_by": "court", "challenge": "none", "ruling": "other", "rules_cited": [], "court_finding_quote": "The judge disclosed using ChatGPT to assist with technical calculations, not as independent evidence.", "text_licence": "quoted-short", "summary": "A Dutch judge disclosed using ChatGPT to assist with technical elements in a private dispute. The row is coded as judicial AI use rather than party evidence; no independent evidentiary challenge or sanction was found in the copied decision text.", "sensitive": false, "publication_ban_checked": false, "related_opinion_id": null, "source_url": "https://uitspraken.rechtspraak.nl/details?id=ECLI:NL:RBGEL:2024:4474", "archive_url": "https://web.archive.org/web/20260907221848/https://uitspraken.rechtspraak.nl/details?id=ECLI:NL:RBGEL:2024:4474", "fetched_at": "2026-09-07T00:00:00Z", "verification": "fetched-and-read", "lead_source": [ "Damien Charlotin acknowledged-AI-output lead CSV (CC0)" ], "notes": "Official Rechtspraak URL was identified from the PDF but failed TLS fetch; downloaded lead PDF copy was read. Publication-ban status not confirmed.", "url": "https://safelegalai.com/courts/evidence/nl-rbgel-2024-x-v-y", "notice": "Provided as is, without warranty. Not legal advice. SafeLegalAI (Cognesio LLP) records what courts, regulators, legislatures and vendors' own public pages state; the linked official documents are the record. Names and marks belong to their owners. Anyone named may reply: https://safelegalai.com/report. Full terms: https://safelegalai.com/disclaimer", "page": "https://safelegalai.com/courts/evidence/nl-rbgel-2024-x-v-y" }, { "decision_id": "us-ny-appdiv-2024-matter-gabriel-h", "case_name": "Matter of Gabriel H.", "court": "New York Appellate Division, Fourth Department", "country": "US", "jurisdiction": "us-state", "state": "NY", "date": "2024-07-03", "citation": "2024 NY Slip Op 03588", "docket_number": "187 CAF 23-01211", "evidence_type": "video", "ai_role": "alleged-deepfake", "offered_by": "petitioner", "challenge": "authenticity", "ruling": "admitted", "rules_cited": [ "People v. Patterson", "Guide to NY Evid rule 9.05" ], "court_finding_quote": "We agree with the court that the videos were sufficiently authenticated and that any alleged uncertainty went to the weight.", "text_licence": "public-domain", "summary": "The Fourth Department affirmed admission of hacked surveillance videos in related child-protection proceedings, relying on FBI copying testimony, law-enforcement identification, matching room details, timestamps, and lack of visible tampering. The court treated deepfake doubts as weight issues; the Court of Appeals later disagreed in Matter of M.S.", "sensitive": true, "publication_ban_checked": true, "related_opinion_id": "ny-appdiv-2024-matter-of-gabriel-h", "source_url": "https://www.nycourts.gov/reporter/3dseries/2024/2024_03588.htm", "archive_url": "http://web.archive.org/web/20240805013155/https://nycourts.gov/reporter/3dseries/2024/2024_03588.htm", "fetched_at": "2026-09-07T00:00:00Z", "verification": "fetched-and-read", "lead_source": [ "Damien Charlotin deepfakes lead CSV (CC0)", "User brief named-case list" ], "notes": null, "url": "https://safelegalai.com/courts/evidence/us-ny-appdiv-2024-matter-gabriel-h", "notice": "Provided as is, without warranty. Not legal advice. SafeLegalAI (Cognesio LLP) records what courts, regulators, legislatures and vendors' own public pages state; the linked official documents are the record. Names and marks belong to their owners. Anyone named may reply: https://safelegalai.com/report. Full terms: https://safelegalai.com/disclaimer", "page": "https://safelegalai.com/courts/evidence/us-ny-appdiv-2024-matter-gabriel-h" }, { "decision_id": "ca-bccrt-2024-yang-v-gibbs", "case_name": "Yang v. Gibbs", "court": "British Columbia Civil Resolution Tribunal", "country": "CA", "jurisdiction": "canada", "state": "BC", "date": "2024-06-27", "citation": "2024 BCCRT 613", "docket_number": "SC-2023-005421", "evidence_type": "chat-output", "ai_role": "acknowledged-ai-output", "offered_by": "applicant", "challenge": "weight", "ruling": "no-weight", "rules_cited": [], "court_finding_quote": "Information provided by ChatGPT is unreliable at best.", "text_licence": "quoted-short", "summary": "Applicants submitted ChatGPT evidence in a fencing dispute. The CRT noted warnings about AI accuracy and prior decisions declining to rely on ChatGPT research, gave the generated material no weight, and decided the claim using ordinary evidence about the contract and work.", "sensitive": false, "publication_ban_checked": false, "related_opinion_id": null, "source_url": "https://www.damiencharlotin.com/documents/569/Yang_v._Gibbs_Canada_24_June_2024.pdf", "archive_url": "https://web.archive.org/web/20260817232955/https://websitedc.s3.amazonaws.com/documents/Yang_v._Gibbs_Canada_24_June_2024.pdf", "fetched_at": "2026-09-07T00:00:00Z", "verification": "fetched-and-read", "lead_source": [ "Damien Charlotin acknowledged-AI-output lead CSV (CC0)" ], "notes": "Official CRT item URL not located; downloaded lead PDF copy was read and publication-ban status not confirmed.", "url": "https://safelegalai.com/courts/evidence/ca-bccrt-2024-yang-v-gibbs", "notice": "Provided as is, without warranty. Not legal advice. SafeLegalAI (Cognesio LLP) records what courts, regulators, legislatures and vendors' own public pages state; the linked official documents are the record. Names and marks belong to their owners. Anyone named may reply: https://safelegalai.com/report. Full terms: https://safelegalai.com/disclaimer", "page": "https://safelegalai.com/courts/evidence/ca-bccrt-2024-yang-v-gibbs" }, { "decision_id": "us-ca11-2024-snell-v-united-specialty", "case_name": "Snell v. United Specialty Insurance Company", "court": "U.S. Court of Appeals for the Eleventh Circuit", "country": "US", "jurisdiction": "us-federal", "state": "AL", "date": "2024-05-28", "citation": null, "docket_number": "22-12581", "evidence_type": "chat-output", "ai_role": "acknowledged-ai-output", "offered_by": "court", "challenge": "none", "ruling": "other", "rules_cited": [ "Alabama insurance law" ], "court_finding_quote": "I didn’t want to fall into the trap of embracing ChatGPT’s definition too readily.", "text_licence": "public-domain", "summary": "Judge Newsom’s concurrence explored whether large language models could inform ordinary meaning in insurance interpretation, while cautioning against uncritical reliance. The panel decided the coverage dispute through conventional legal analysis; the AI discussion was methodological and not party evidence.", "sensitive": false, "publication_ban_checked": true, "related_opinion_id": "ca11-2024-snell-v-united-specialty", "source_url": "https://media.ca11.uscourts.gov/opinions/pub/files/202212581.pdf", "archive_url": "https://web.archive.org/web/20260825211740/https://media.ca11.uscourts.gov/opinions/pub/files/202212581.pdf", "fetched_at": "2026-09-07T00:00:00Z", "verification": "fetched-and-read", "lead_source": [ "Damien Charlotin acknowledged-AI-output lead CSV (CC0)" ], "notes": null, "url": "https://safelegalai.com/courts/evidence/us-ca11-2024-snell-v-united-specialty", "notice": "Provided as is, without warranty. Not legal advice. SafeLegalAI (Cognesio LLP) records what courts, regulators, legislatures and vendors' own public pages state; the linked official documents are the record. Names and marks belong to their owners. Anyone named may reply: https://safelegalai.com/report. Full terms: https://safelegalai.com/disclaimer", "page": "https://safelegalai.com/courts/evidence/us-ca11-2024-snell-v-united-specialty" }, { "decision_id": "in-mani-hc-2024-hussain-v-state-manipur", "case_name": "Hussain v. State of Manipur", "court": "High Court of Manipur", "country": "IN", "jurisdiction": "india", "state": "Manipur", "date": "2024-05-23", "citation": null, "docket_number": null, "evidence_type": "chat-output", "ai_role": "acknowledged-ai-output", "offered_by": "court", "challenge": "none", "ruling": "other", "rules_cited": [], "court_finding_quote": "The court disclosed using Google and ChatGPT 3.5 to research background details counsel had not covered.", "text_licence": "quoted-short", "summary": "The Manipur High Court disclosed using Google and ChatGPT 3.5 for background research on factual details in a bail or criminal matter. The AI material was framed as supplemental context rather than evidence controlling the result.", "sensitive": true, "publication_ban_checked": false, "related_opinion_id": null, "source_url": "https://www.damiencharlotin.com/documents/432/Hussain_vs_State_of_Manipur_Manpur_India_May_23_2025.pdf", "archive_url": "https://web.archive.org/web/20260817232955/https://websitedc.s3.amazonaws.com/documents/Hussain_vs_State_of_Manipur_Manpur_India_May_23_2025.pdf", "fetched_at": "2026-09-07T00:00:00Z", "verification": "fetched-and-read", "lead_source": [ "Damien Charlotin acknowledged-AI-output lead CSV (CC0)" ], "notes": "Downloaded lead PDF copy read; no official Indian court copy or publication-ban status confirmed.", "url": "https://safelegalai.com/courts/evidence/in-mani-hc-2024-hussain-v-state-manipur", "notice": "Provided as is, without warranty. Not legal advice. SafeLegalAI (Cognesio LLP) records what courts, regulators, legislatures and vendors' own public pages state; the linked official documents are the record. Names and marks belong to their owners. Anyone named may reply: https://safelegalai.com/report. Full terms: https://safelegalai.com/disclaimer", "page": "https://safelegalai.com/courts/evidence/in-mani-hc-2024-hussain-v-state-manipur" }, { "decision_id": "ca-qccs-2024-azubuike-c-r", "case_name": "Azubuike c. R.", "court": "Superior Court of Québec", "country": "CA", "jurisdiction": "canada", "state": "QC", "date": "2024-05-09", "citation": "2024 QCCS 1654", "docket_number": "500-01-207497-204", "evidence_type": "audio", "ai_role": "voice-clone", "offered_by": "prosecution", "challenge": "expert-required", "ruling": "excluded", "rules_cited": [ "Mohan expert evidence test", "Canada Evidence Act s.31.1" ], "court_finding_quote": "The proposed voice-cloning expert was not qualified; his report and testimony were excluded.", "text_licence": "quoted-short", "summary": "The accused alleged eight customer-service recordings were voice clones and offered a proposed expert. After a voir dire, the court found the expert unqualified and excluded his report and testimony, while leaving the recordings’ ultimate authenticity to be addressed through other evidence.", "sensitive": false, "publication_ban_checked": false, "related_opinion_id": null, "source_url": "https://www.canlii.org/fr/qc/qccs/doc/2024/2024qccs1654/2024qccs1654.html", "archive_url": null, "fetched_at": "2026-09-07T00:00:00Z", "verification": "fetched-and-read", "lead_source": [ "Damien Charlotin deepfakes lead CSV (CC0)", "User brief named-case list" ], "notes": "CanLII page could not be independently read by script; downloaded lead PDF copy was read.", "url": "https://safelegalai.com/courts/evidence/ca-qccs-2024-azubuike-c-r", "notice": "Provided as is, without warranty. Not legal advice. SafeLegalAI (Cognesio LLP) records what courts, regulators, legislatures and vendors' own public pages state; the linked official documents are the record. Names and marks belong to their owners. Anyone named may reply: https://safelegalai.com/report. Full terms: https://safelegalai.com/disclaimer", "page": "https://safelegalai.com/courts/evidence/ca-qccs-2024-azubuike-c-r" }, { "decision_id": "us-wa-super-2024-state-v-puloka", "case_name": "State v. Puloka", "court": "King County Superior Court, Washington", "country": "US", "jurisdiction": "us-state", "state": "WA", "date": "2024-04-01", "citation": null, "docket_number": "21-1-04851-2 KNT", "evidence_type": "video", "ai_role": "ai-enhanced", "offered_by": "defence", "challenge": "admissibility", "ruling": "excluded", "rules_cited": [ "Frye", "ER 702", "ER 403" ], "court_finding_quote": "The video evidence produced by the Topaz Video AI enhancement model does not show what happened but uses opaque methods to represent what the AI model thinks should be there.", "text_licence": "public-domain", "summary": "The defense sought to admit AI-enhanced video processed with Topaz Labs AI. After a Frye hearing, the court found the machine-learning method opaque, not generally accepted in forensic video analysis, potentially misleading, and unduly prejudicial, excluding the enhanced videos while leaving original video available.", "sensitive": false, "publication_ban_checked": false, "related_opinion_id": "wash-super-2024-state-v-puloka", "source_url": "https://news.bloomberglaw.com/litigation/judges-ai-video-rejection-evokes-broader-tech-evidence-issues", "archive_url": "http://web.archive.org/web/20260907213313/https://news.bloomberglaw.com/litigation/judges-ai-video-rejection-evokes-broader-tech-evidence-issues", "fetched_at": "2026-09-07T00:00:00Z", "verification": "link-only", "lead_source": [ "User brief named-case list" ], "notes": "No official court copy found; OCR was made from a circulated PDF. Source URL is a news-hosted discussion, not the Thomson Reuters graphics URL; publication-ban status not confirmed.", "url": "https://safelegalai.com/courts/evidence/us-wa-super-2024-state-v-puloka", "notice": "Provided as is, without warranty. Not legal advice. SafeLegalAI (Cognesio LLP) records what courts, regulators, legislatures and vendors' own public pages state; the linked official documents are the record. Names and marks belong to their owners. Anyone named may reply: https://safelegalai.com/report. Full terms: https://safelegalai.com/disclaimer", "page": "https://safelegalai.com/courts/evidence/us-wa-super-2024-state-v-puloka" }, { "decision_id": "us-sdny-2024-jg-v-nyc-doe", "case_name": "J.G. v. New York City Department of Education", "court": "U.S. District Court for the Southern District of New York", "country": "US", "jurisdiction": "us-federal", "state": "NY", "date": "2024-02-22", "citation": null, "docket_number": "23 Civ. 959", "evidence_type": "chat-output", "ai_role": "acknowledged-ai-output", "offered_by": "plaintiff", "challenge": "weight", "ruling": "no-weight", "rules_cited": [ "IDEA", "20 U.S.C. § 1415(i)(3)", "Lodestar method" ], "court_finding_quote": "The court found the ChatGPT-4 submission utterly and unusually unpersuasive as support for prevailing market rates.", "text_licence": "public-domain", "summary": "Plaintiff’s counsel used GPT-4 as a cross-check for requested IDEA fee rates. Judge Engelmayer rejected the AI output as opaque and unreliable evidence of prevailing market rates, reduced the fee award, and warned against using ChatGPT to support legal fee petitions.", "sensitive": true, "publication_ban_checked": true, "related_opinion_id": "sdny-2024-jg-v-nyc-doe", "source_url": "https://www.damiencharlotin.com/documents/248/JG_v_NYC_Department_of_Education_US_DC_SDNY_February_22_2024.pdf", "archive_url": null, "fetched_at": "2026-09-07T00:00:00Z", "verification": "fetched-and-read", "lead_source": [ "Damien Charlotin acknowledged-AI-output lead CSV (CC0)" ], "notes": null, "url": "https://safelegalai.com/courts/evidence/us-sdny-2024-jg-v-nyc-doe", "notice": "Provided as is, without warranty. Not legal advice. SafeLegalAI (Cognesio LLP) records what courts, regulators, legislatures and vendors' own public pages state; the linked official documents are the record. Names and marks belong to their owners. Anyone named may reply: https://safelegalai.com/report. Full terms: https://safelegalai.com/disclaimer", "page": "https://safelegalai.com/courts/evidence/us-sdny-2024-jg-v-nyc-doe" }, { "decision_id": "ca-abkb-2024-r-v-md", "case_name": "R v MD", "court": "Court of King's Bench of Alberta", "country": "CA", "jurisdiction": "canada", "state": "AB", "date": "2024-02-22", "citation": "2024 ABKB 104", "docket_number": "220314306Q1", "evidence_type": "audio", "ai_role": "voice-clone", "offered_by": "prosecution", "challenge": "authenticity", "ruling": "admitted", "rules_cited": [ "Canada Evidence Act s.31.1" ], "court_finding_quote": "The judge found no evidence of tampering or AI fabrication and accepted the male voice as the accused.", "text_licence": "quoted-short", "summary": "The complainant offered a cellphone recording of an inculpatory porch conversation. Defence suggested the male voice had been cloned. After a voir dire, the judge relied on the complainant’s evidence, voice comparison, and contextual fit, rejected AI fabrication, admitted the recording, and convicted.", "sensitive": true, "publication_ban_checked": false, "related_opinion_id": null, "source_url": "https://www.canlii.org/en/ab/abkb/doc/2024/2024abkb104/2024abkb104.html", "archive_url": null, "fetched_at": "2026-09-07T00:00:00Z", "verification": "fetched-and-read", "lead_source": [ "Damien Charlotin deepfakes lead CSV (CC0)", "User brief named-case list" ], "notes": "CanLII page could not be independently read by script; downloaded lead PDF copy was read; publication-ban status not confirmed from the live page.", "url": "https://safelegalai.com/courts/evidence/ca-abkb-2024-r-v-md", "notice": "Provided as is, without warranty. Not legal advice. SafeLegalAI (Cognesio LLP) records what courts, regulators, legislatures and vendors' own public pages state; the linked official documents are the record. Names and marks belong to their owners. Anyone named may reply: https://safelegalai.com/report. Full terms: https://safelegalai.com/disclaimer", "page": "https://safelegalai.com/courts/evidence/ca-abkb-2024-r-v-md" }, { "decision_id": "au-vsc-2024-dpp-v-khan", "case_name": "DPP v. Khan", "court": "Supreme Court of Victoria", "country": "AU", "jurisdiction": "australia", "state": "VIC", "date": "2024-02-07", "citation": null, "docket_number": null, "evidence_type": "witness-statement", "ai_role": "ai-generated-submitted", "offered_by": "defence", "challenge": "weight", "ruling": "no-weight", "rules_cited": [], "court_finding_quote": "It is clearly inappropriate that personal references used in sentencing are generated by large language models.", "text_licence": "quoted-short", "summary": "Defence counsel tendered a character reference that appeared translated or generated with a large language model. The sentencing judge found the absence of evidence about generation made weight difficult to assess, treated LLM-generated references as inappropriate, and ignored the reference.", "sensitive": false, "publication_ban_checked": false, "related_opinion_id": null, "source_url": "https://www.damiencharlotin.com/documents/464/DPP_v._Khan_Australia_7_February_2024.pdf", "archive_url": "https://web.archive.org/web/20260817232955/https://websitedc.s3.amazonaws.com/documents/DPP_v._Khan_Australia_7_February_2024.pdf", "fetched_at": "2026-09-07T00:00:00Z", "verification": "fetched-and-read", "lead_source": [ "Damien Charlotin acknowledged-AI-output lead CSV (CC0)" ], "notes": "Downloaded lead PDF copy read; publication-ban status not confirmed.", "url": "https://safelegalai.com/courts/evidence/au-vsc-2024-dpp-v-khan", "notice": "Provided as is, without warranty. Not legal advice. SafeLegalAI (Cognesio LLP) records what courts, regulators, legislatures and vendors' own public pages state; the linked official documents are the record. Names and marks belong to their owners. Anyone named may reply: https://safelegalai.com/report. Full terms: https://safelegalai.com/disclaimer", "page": "https://safelegalai.com/courts/evidence/au-vsc-2024-dpp-v-khan" }, { "decision_id": "us-pa-super-2023-commonwealth-v-spone", "case_name": "Commonwealth v. Spone", "court": "Superior Court of Pennsylvania", "country": "US", "jurisdiction": "us-state", "state": "PA", "date": "2023-11-14", "citation": null, "docket_number": "1623 EDA 2022", "evidence_type": "mixed", "ai_role": "alleged-deepfake", "offered_by": "prosecution", "challenge": "authenticity", "ruling": "admitted", "rules_cited": [ "Pa.R.E. 901" ], "court_finding_quote": "Witnesses testified they believed some photos or videos were manufactured, but no judicial finding of synthesis was made.", "text_licence": "public-domain", "summary": "Anonymous texts containing images and videos of minors were traced through Pinger records, IP evidence, and phone data to Spone. Some witnesses believed media had been manufactured and public reports used deepfake language, but the courts did not resolve AI generation and affirmed the convictions.", "sensitive": true, "publication_ban_checked": true, "related_opinion_id": "pa-super-2023-commonwealth-v-spone", "source_url": "https://www.damiencharlotin.com/documents/3001/2023-1623-eda-2022.pdf", "archive_url": "https://web.archive.org/web/20260907220721/https://websitedc.s3.amazonaws.com/documents/2023-1623-eda-2022.pdf", "fetched_at": "2026-09-07T00:00:00Z", "verification": "fetched-and-read", "lead_source": [ "Damien Charlotin deepfakes lead CSV (CC0)", "User brief named-case list" ], "notes": null, "url": "https://safelegalai.com/courts/evidence/us-pa-super-2023-commonwealth-v-spone", "notice": "Provided as is, without warranty. Not legal advice. SafeLegalAI (Cognesio LLP) records what courts, regulators, legislatures and vendors' own public pages state; the linked official documents are the record. Names and marks belong to their owners. Anyone named may reply: https://safelegalai.com/report. Full terms: https://safelegalai.com/disclaimer", "page": "https://safelegalai.com/courts/evidence/us-pa-super-2023-commonwealth-v-spone" }, { "decision_id": "us-dsc-2023-pegnatori-v-pure-sports", "case_name": "Pegnatori v. Pure Sports Technology", "court": "U.S. District Court for the District of South Carolina", "country": "US", "jurisdiction": "us-federal", "state": "SC", "date": "2023-10-11", "citation": null, "docket_number": "2:23-cv-01424", "evidence_type": "chat-output", "ai_role": "acknowledged-ai-output", "offered_by": "defendant", "challenge": "weight", "ruling": "no-weight", "rules_cited": [ "Patent claim construction", "Fed. R. Evid. 802" ], "court_finding_quote": "The court would be taking its eye off the ball if it applied the ChatGPT definition in its review of extrinsic evidence.", "text_licence": "public-domain", "summary": "Pure Sports offered a ChatGPT definition of foam in a patent dispute. The court rejected it because the patent’s filing date predated ChatGPT and claim construction asks what a skilled artisan would understand then, not a contemporary chatbot answer.", "sensitive": false, "publication_ban_checked": true, "related_opinion_id": "dsc-2023-pegnatori-v-pure-sports", "source_url": "https://www.damiencharlotin.com/documents/582/Pegnatori_v._Pure_Sports_Technology_USA_11_October_2023.pdf", "archive_url": "https://web.archive.org/web/20260817232955/https://websitedc.s3.amazonaws.com/documents/Pegnatori_v._Pure_Sports_Technology_USA_11_October_2023.pdf", "fetched_at": "2026-09-07T00:00:00Z", "verification": "fetched-and-read", "lead_source": [ "Damien Charlotin acknowledged-AI-output lead CSV (CC0)" ], "notes": null, "url": "https://safelegalai.com/courts/evidence/us-dsc-2023-pegnatori-v-pure-sports", "notice": "Provided as is, without warranty. Not legal advice. SafeLegalAI (Cognesio LLP) records what courts, regulators, legislatures and vendors' own public pages state; the linked official documents are the record. Names and marks belong to their owners. Anyone named may reply: https://safelegalai.com/report. Full terms: https://safelegalai.com/disclaimer", "page": "https://safelegalai.com/courts/evidence/us-dsc-2023-pegnatori-v-pure-sports" }, { "decision_id": "in-delhi-hc-2023-louboutin-v-shoe-boutique", "case_name": "Louboutin v. The Shoe Boutique", "court": "Delhi High Court", "country": "IN", "jurisdiction": "india", "state": "Delhi", "date": "2023-08-22", "citation": null, "docket_number": null, "evidence_type": "chat-output", "ai_role": "acknowledged-ai-output", "offered_by": "plaintiff", "challenge": "weight", "ruling": "no-weight", "rules_cited": [], "court_finding_quote": "AI cannot substitute either the human intelligence or the humane element in the adjudicatory process.", "text_licence": "quoted-short", "summary": "Plaintiff’s counsel cited ChatGPT responses to support reputation claims, and the court ran its own illustrative prompts. The Delhi High Court said chatbot answers vary with prompts, may be inaccurate or fictional, and cannot decide legal or factual issues.", "sensitive": false, "publication_ban_checked": false, "related_opinion_id": null, "source_url": "https://www.damiencharlotin.com/documents/433/Louboutin_v._the_Shoe_Boutique_Delhi_High_Court_August_22_2023.pdf", "archive_url": null, "fetched_at": "2026-09-07T00:00:00Z", "verification": "fetched-and-read", "lead_source": [ "Damien Charlotin acknowledged-AI-output lead CSV (CC0)" ], "notes": "Downloaded lead PDF copy read; no official Indian court source or publication-ban status confirmed.", "url": "https://safelegalai.com/courts/evidence/in-delhi-hc-2023-louboutin-v-shoe-boutique", "notice": "Provided as is, without warranty. Not legal advice. SafeLegalAI (Cognesio LLP) records what courts, regulators, legislatures and vendors' own public pages state; the linked official documents are the record. Names and marks belong to their owners. Anyone named may reply: https://safelegalai.com/report. Full terms: https://safelegalai.com/disclaimer", "page": "https://safelegalai.com/courts/evidence/in-delhi-hc-2023-louboutin-v-shoe-boutique" }, { "decision_id": "uk-nihc-2023-santander-v-carlin", "case_name": "Santander v. Carlin", "court": "High Court of Justice in Northern Ireland", "country": "GB", "jurisdiction": "uk", "state": null, "date": "2023-06-16", "citation": "[2023] NICh 5", "docket_number": null, "evidence_type": "chat-output", "ai_role": "acknowledged-ai-output", "offered_by": "defendant", "challenge": "weight", "ruling": "no-weight", "rules_cited": [], "court_finding_quote": "ChatGPT general comments did not provide evidence or a legal answer to the bank's claim.", "text_licence": "quoted-short", "summary": "A self-represented litigant relied on ChatGPT-generated general comments about the opposing side’s case. The Northern Ireland High Court treated the material as unhelpful and gave it no evidentiary weight, resolving the mortgage and possession dispute on ordinary proof and law.", "sensitive": false, "publication_ban_checked": true, "related_opinion_id": null, "source_url": "https://www.judiciaryni.uk/judicial-decisions/2023-nich-5", "archive_url": "https://web.archive.org/web/20260907213642/https://www.judiciaryni.uk/judicial-decisions/2023-nich-5", "fetched_at": "2026-09-07T00:00:00Z", "verification": "page-read-no-document", "lead_source": [ "Damien Charlotin acknowledged-AI-output lead CSV (CC0)" ], "notes": "Official Judiciary NI decision page was fetched; the document link was not extracted in this session. No publication-ban issue identified on the page.", "url": "https://safelegalai.com/courts/evidence/uk-nihc-2023-santander-v-carlin", "notice": "Provided as is, without warranty. Not legal advice. SafeLegalAI (Cognesio LLP) records what courts, regulators, legislatures and vendors' own public pages state; the linked official documents are the record. Names and marks belong to their owners. Anyone named may reply: https://safelegalai.com/report. Full terms: https://safelegalai.com/disclaimer", "page": "https://safelegalai.com/courts/evidence/uk-nihc-2023-santander-v-carlin" }, { "decision_id": "us-ndcal-2023-valenti-v-dfinity", "case_name": "Valenti v. Dfinity USA Research LLC", "court": "U.S. District Court for the Northern District of California", "country": "US", "jurisdiction": "us-federal", "state": "CA", "date": "2023-05-08", "citation": null, "docket_number": "21-cv-06118", "evidence_type": "video", "ai_role": "deepfake-defence", "offered_by": "plaintiff", "challenge": "admissibility", "ruling": "excluded", "rules_cited": [ "Fed. R. Civ. P. 23", "PSLRA" ], "court_finding_quote": "The court denied leave to file supplemental expert material asserting that controversial CryptoLeaks videos might be deepfakes.", "text_licence": "public-domain", "summary": "FNF sought to file supplemental expert evidence suggesting CryptoLeaks videos of counsel could be deepfakes. The court denied the request in an order removing FNF as lead counsel on class-representation grounds, leaving the authenticity of the videos unresolved and excluding the supplemental AI evidence.", "sensitive": false, "publication_ban_checked": true, "related_opinion_id": "ndcal-2023-valenti-v-dfinity", "source_url": "https://www.damiencharlotin.com/documents/2963/0_3.pdf", "archive_url": null, "fetched_at": "2026-09-07T00:00:00Z", "verification": "fetched-and-read", "lead_source": [ "Damien Charlotin deepfakes lead CSV (CC0)", "User brief named-case list" ], "notes": null, "url": "https://safelegalai.com/courts/evidence/us-ndcal-2023-valenti-v-dfinity", "notice": "Provided as is, without warranty. Not legal advice. SafeLegalAI (Cognesio LLP) records what courts, regulators, legislatures and vendors' own public pages state; the linked official documents are the record. Names and marks belong to their owners. Anyone named may reply: https://safelegalai.com/report. Full terms: https://safelegalai.com/disclaimer", "page": "https://safelegalai.com/courts/evidence/us-ndcal-2023-valenti-v-dfinity" }, { "decision_id": "us-sdny-2023-freedman-v-cyrulnik", "case_name": "Freedman Normand Friedland LLP v. Cyrulnik", "court": "U.S. District Court for the Southern District of New York", "country": "US", "jurisdiction": "us-federal", "state": "NY", "date": "2023-04-05", "citation": null, "docket_number": "1:21-cv-01746", "evidence_type": "expert-report", "ai_role": "deepfake-defence", "offered_by": "plaintiff", "challenge": "admissibility", "ruling": "not-adjudicated", "rules_cited": [ "Fed. R. Evid. 702", "Fed. R. Civ. P. 37" ], "court_finding_quote": null, "text_licence": "public-domain", "summary": "The brief identified an April 2023 order involving proposed deepfake expert material. Public searches located docket summaries but no court-hosted ECF 380 or full order. The row records the lead as unconfirmed and should be revisited with PACER or another primary copy.", "sensitive": false, "publication_ban_checked": false, "related_opinion_id": "sdny-2023-freedman-v-cyrulnik", "source_url": "https://www.courtlistener.com/docket/59894977/freedman-normand-friedland-llp-v-cyrulnik/", "archive_url": null, "fetched_at": "2026-09-07T00:00:00Z", "verification": "link-only", "lead_source": [ "User brief named-case list" ], "notes": "No publicly fetchable primary order located; CourtListener docket page used only as locator, not a read opinion. Publication-ban status not confirmed.", "url": "https://safelegalai.com/courts/evidence/us-sdny-2023-freedman-v-cyrulnik", "notice": "Provided as is, without warranty. Not legal advice. SafeLegalAI (Cognesio LLP) records what courts, regulators, legislatures and vendors' own public pages state; the linked official documents are the record. Names and marks belong to their owners. Anyone named may reply: https://safelegalai.com/report. Full terms: https://safelegalai.com/disclaimer", "page": "https://safelegalai.com/courts/evidence/us-sdny-2023-freedman-v-cyrulnik" }, { "decision_id": "us-ca-super-2023-huang-v-tesla", "case_name": "Huang v. Tesla, Inc.", "court": "Superior Court of California, Santa Clara County", "country": "US", "jurisdiction": "us-state", "state": "CA", "date": "2023-04-05", "citation": null, "docket_number": "19CV346663", "evidence_type": "video", "ai_role": "deepfake-defence", "offered_by": "defendant", "challenge": "authenticity", "ruling": "authentication-ordered", "rules_cited": [ "Cal. Civ. Proc. Code § 2031.310" ], "court_finding_quote": "Tesla cannot hide behind the potential for recorded statements being a deep fake to avoid taking a position.", "text_licence": "public-domain", "summary": "In discovery over Elon Musk’s public statements, Tesla resisted authenticating recordings by invoking the general possibility of deepfakes. The court rejected that position, reasoning Musk could confirm whether he appeared and spoke, and required Tesla to provide fuller responses or present the recordings to Musk.", "sensitive": false, "publication_ban_checked": false, "related_opinion_id": "cal-super-2023-huang-v-tesla", "source_url": "https://cdn.arstechnica.net/wp-content/uploads/2023/04/tesla-tentative-ruling.pdf", "archive_url": null, "fetched_at": "2026-09-07T00:00:00Z", "verification": "link-only", "lead_source": [ "User brief named-case list" ], "notes": "Tentative ruling copy was fetched from Ars Technica CDN; no court-hosted copy located and publication-ban status not confirmed.", "url": "https://safelegalai.com/courts/evidence/us-ca-super-2023-huang-v-tesla", "notice": "Provided as is, without warranty. Not legal advice. SafeLegalAI (Cognesio LLP) records what courts, regulators, legislatures and vendors' own public pages state; the linked official documents are the record. Names and marks belong to their owners. Anyone named may reply: https://safelegalai.com/report. Full terms: https://safelegalai.com/disclaimer", "page": "https://safelegalai.com/courts/evidence/us-ca-super-2023-huang-v-tesla" }, { "decision_id": "in-punjab-hc-2023-jaswinder-singh-v-state", "case_name": "Jaswinder Singh v. State of Punjab", "court": "Punjab and Haryana High Court", "country": "IN", "jurisdiction": "india", "state": "Punjab", "date": "2023-03-27", "citation": null, "docket_number": null, "evidence_type": "chat-output", "ai_role": "acknowledged-ai-output", "offered_by": "court", "challenge": "none", "ruling": "other", "rules_cited": [], "court_finding_quote": "The ChatGPT reference was only intended to present a broader picture on bail jurisprudence.", "text_licence": "quoted-short", "summary": "The High Court used ChatGPT to ask about bail when assault is laced with cruelty, then expressly limited the reference. It stated the chatbot answer was not an opinion on the merits and should not be used by the trial court.", "sensitive": true, "publication_ban_checked": false, "related_opinion_id": null, "source_url": "https://www.damiencharlotin.com/documents/431/Jaswinder_Singh_Punjab_India_March_27_2023.pdf", "archive_url": "https://web.archive.org/web/20260817232955/https://websitedc.s3.amazonaws.com/documents/Jaswinder_Singh_Punjab_India_March_27_2023.pdf", "fetched_at": "2026-09-07T00:00:00Z", "verification": "fetched-and-read", "lead_source": [ "Damien Charlotin acknowledged-AI-output lead CSV (CC0)" ], "notes": "Downloaded lead PDF copy read; no official Indian court source or publication-ban status confirmed.", "url": "https://safelegalai.com/courts/evidence/in-punjab-hc-2023-jaswinder-singh-v-state", "notice": "Provided as is, without warranty. Not legal advice. SafeLegalAI (Cognesio LLP) records what courts, regulators, legislatures and vendors' own public pages state; the linked official documents are the record. Names and marks belong to their owners. Anyone named may reply: https://safelegalai.com/report. Full terms: https://safelegalai.com/disclaimer", "page": "https://safelegalai.com/courts/evidence/in-punjab-hc-2023-jaswinder-singh-v-state" }, { "decision_id": "co-ct-2023-salvador-espitia-chavez", "case_name": "Salvador Espitia Chavez v. Salud Total", "court": "Constitutional Court of Colombia", "country": "CO", "jurisdiction": "other", "state": null, "date": "2023-01-01", "citation": null, "docket_number": null, "evidence_type": "document", "ai_role": "acknowledged-ai-output", "offered_by": "court", "challenge": "none", "ruling": "other", "rules_cited": [], "court_finding_quote": "The court clerk used ChatGPT prompts for legal background and drafting, not party-proffered evidence.", "text_licence": "quoted-short", "summary": "A Colombian court decision disclosed or was reported to include ChatGPT-assisted drafting about health-care exemptions for a minor with disabilities. The row captures judicial use of AI in preparing legal reasoning rather than a contested evidentiary exhibit.", "sensitive": true, "publication_ban_checked": false, "related_opinion_id": null, "source_url": "https://www.damiencharlotin.com/documents/435/Salvador_Espitia_Chavez_v._Salud_Total_Colombia_January_1_2023.pdf", "archive_url": "https://web.archive.org/web/20260907222120/https://websitedc.s3.amazonaws.com/documents/Salvador_Espitia_Chavez_v._Salud_Total_Colombia_January_1_2023.pdf", "fetched_at": "2026-09-07T00:00:00Z", "verification": "fetched-and-read", "lead_source": [ "Damien Charlotin acknowledged-AI-output lead CSV (CC0)" ], "notes": "Downloaded lead PDF copy read; no official Colombian source or publication-ban status confirmed.", "url": "https://safelegalai.com/courts/evidence/co-ct-2023-salvador-espitia-chavez", "notice": "Provided as is, without warranty. Not legal advice. SafeLegalAI (Cognesio LLP) records what courts, regulators, legislatures and vendors' own public pages state; the linked official documents are the record. Names and marks belong to their owners. Anyone named may reply: https://safelegalai.com/report. Full terms: https://safelegalai.com/disclaimer", "page": "https://safelegalai.com/courts/evidence/co-ct-2023-salvador-espitia-chavez" }, { "decision_id": "us-ddc-2022-united-states-v-reffitt", "case_name": "United States v. Reffitt", "court": "U.S. District Court for the District of Columbia", "country": "US", "jurisdiction": "us-federal", "state": "DC", "date": "2022-04-10", "citation": "602 F. Supp. 3d 85", "docket_number": "1:21-cr-00032", "evidence_type": "video", "ai_role": "other", "offered_by": "prosecution", "challenge": "none", "ruling": "admitted", "rules_cited": [ "Fed. R. Evid. 401", "Fed. R. Evid. 403" ], "court_finding_quote": null, "text_licence": "public-domain", "summary": "The Reffitt materials reviewed on GovInfo concern conventional January 6 video and audio evidence, not AI fabrication or deepfake authentication. The row is retained because the case was named in the collection brief; no AI-specific evidentiary ruling was confirmed in the public document read.", "sensitive": false, "publication_ban_checked": true, "related_opinion_id": "ddc-2022-united-states-v-reffitt", "source_url": "https://www.govinfo.gov/content/pkg/USCOURTS-dcd-1_21-cr-00032/pdf/USCOURTS-dcd-1_21-cr-00032-0.pdf", "archive_url": null, "fetched_at": "2026-09-07T00:00:00Z", "verification": "fetched-and-read", "lead_source": [ "User brief named-case list" ], "notes": "Zero-finding row: no AI/deepfake issue located in the GovInfo document read.", "url": "https://safelegalai.com/courts/evidence/us-ddc-2022-united-states-v-reffitt", "notice": "Provided as is, without warranty. Not legal advice. SafeLegalAI (Cognesio LLP) records what courts, regulators, legislatures and vendors' own public pages state; the linked official documents are the record. Names and marks belong to their owners. Anyone named may reply: https://safelegalai.com/report. Full terms: https://safelegalai.com/disclaimer", "page": "https://safelegalai.com/courts/evidence/us-ddc-2022-united-states-v-reffitt" } ], "instruments": [ { "instrument_id": "ca-evidence-act-section-31-1", "title": "Canada Evidence Act section 31.1", "body": "Electronic-document authentication rule", "country": "CA", "jurisdiction": "canada", "state": null, "type": "rule", "status": "in-force", "status_date": "2026-09-07", "status_quote": "Any person seeking to admit an electronic document as evidence has the burden of proving its authenticity.", "effective_date": null, "next_event": null, "operative_text": "Any person seeking to admit an electronic document as evidence has the burden of proving its authenticity by evidence capable of supporting a finding that the electronic document is that which it is purported to be.", "text_licence": "quoted-short", "summary": "Canada Evidence Act section 31.1 is not AI-specific, but it supplies the authentication burden repeatedly applied in Canadian deepfake and digital-evidence cases. It requires proof capable of supporting a finding that an electronic document is what it purports to be.", "source_url": "https://laws-lois.justice.gc.ca/eng/acts/C-5/section-31.1.html", "archive_url": "https://web.archive.org/web/20260907222304/https://laws-lois.justice.gc.ca/eng/acts/C-5/section-31.1.html", "fetched_at": "2026-09-07T00:00:00Z", "verification": "fetched-and-read", "lead_source": [ "Justice Laws website", "User brief" ], "notes": null, "url": "https://safelegalai.com/courts/evidence/rules/ca-evidence-act-section-31-1", "notice": "Provided as is, without warranty. Not legal advice. SafeLegalAI (Cognesio LLP) records what courts, regulators, legislatures and vendors' own public pages state; the linked official documents are the record. Names and marks belong to their owners. Anyone named may reply: https://safelegalai.com/report. Full terms: https://safelegalai.com/disclaimer", "page": "https://safelegalai.com/courts/evidence/rules/ca-evidence-act-section-31-1" }, { "instrument_id": "us-fre-901c-draft", "title": "Draft Federal Rule of Evidence 901(c)", "body": "Deepfake authentication burden-shifting draft", "country": "US", "jurisdiction": "us-federal", "state": null, "type": "proposed-rule", "status": "under-study", "status_date": "2026-05-07", "status_quote": "The Committee has largely coalesced on what a rule amendment might look like.", "effective_date": null, "next_event": "Potential publication for public comment was discussed for the Spring 2026 meeting; no final text was advanced in the checked drafts.", "operative_text": "Rule 901(c). Potentially Fabricated Evidence Created by Generative Artificial Intelligence. (1) Showing Required to Warrant an Inquiry into Fabrication. If a party challenges the authenticity of an item of evidence on the ground that it has been fabricated, in whole or in part, by generative artificial intelligence, the party must present evidence sufficient to support a finding of fabrication to warrant an inquiry by the court. (2) Showing Then Required by the Proponent. If the opponent makes that showing, the item of evidence is admissible only if the proponent demonstrates to the court that it is more likely than not authentic. (3) Applicability. This rule applies to items offered under either Rule 901 or 902. (4) Notice. Unless the court orders otherwise, a party claiming that an item is fabricated in whole or part by generative artificial intelligence must provide reasonable pretrial notice to all opposing parties of the intent to present evidence of fabrication, so that the opposing parties have a reasonable opportunity to respond to that evidence before trial. (5) Definitions. In this rule, generative artificial intelligence means a computer system that emulates the structure and characteristics of input data in order to create derived synthetic content, including images, video, audio, text, and other digital content.", "text_licence": "public-domain", "summary": "Draft Rule 901(c) would require an opponent to make a threshold showing that generative AI fabrication warrants inquiry, then require the proponent to prove authenticity by a preponderance. The May 2026 agenda treats the draft as under study pending survey results and committee action.", "source_url": "https://www.uscourts.gov/sites/default/files/document/2026-05-evidence-rules-agenda-book.pdf", "archive_url": "https://web.archive.org/web/20260907222159/https://www.uscourts.gov/sites/default/files/document/2026-05-evidence-rules-agenda-book.pdf", "fetched_at": "2026-09-07T00:00:00Z", "verification": "fetched-and-read", "lead_source": [ "May 2026 Evidence Rules agenda book" ], "notes": null, "url": "https://safelegalai.com/courts/evidence/rules/us-fre-901c-draft", "notice": "Provided as is, without warranty. Not legal advice. SafeLegalAI (Cognesio LLP) records what courts, regulators, legislatures and vendors' own public pages state; the linked official documents are the record. Names and marks belong to their owners. Anyone named may reply: https://safelegalai.com/report. Full terms: https://safelegalai.com/disclaimer", "page": "https://safelegalai.com/courts/evidence/rules/us-fre-901c-draft" }, { "instrument_id": "us-fre-707-proposed", "title": "Proposed Federal Rule of Evidence 707", "body": "Machine-generated evidence reliability gatekeeping", "country": "US", "jurisdiction": "us-federal", "state": null, "type": "proposed-rule", "status": "not-advanced", "status_date": "2026-05-07", "status_quote": "The Advisory Committee has not come to a formal decision on this issue.", "effective_date": null, "next_event": "No Evidence Rule 707 appears in the August 2026 preliminary draft; monitor future Evidence Rules agenda books.", "operative_text": "Rule 707. Machine-Generated Evidence When machine-generated evidence is offered without an expert witness and would be subject to Rule 702 if testified to by a witness, the court may admit the evidence only if it satisfies the requirements of Rule 702(a)-(d). This rule does not apply to the output of simple scientific instruments.", "text_licence": "public-domain", "summary": "The Advisory Committee published proposed Rule 707 in August 2025 to apply Rule 702 reliability standards when machine-generated evidence substitutes for expert testimony. The May 2026 agenda reflects continued study and controversy rather than advancement to the next rulemaking stage.", "source_url": "https://www.uscourts.gov/sites/default/files/document/2026-05-evidence-rules-agenda-book.pdf", "archive_url": "https://web.archive.org/web/20260907222159/https://www.uscourts.gov/sites/default/files/document/2026-05-evidence-rules-agenda-book.pdf", "fetched_at": "2026-09-07T00:00:00Z", "verification": "fetched-and-read", "lead_source": [ "U.S. Courts pending rules page", "May 2026 Evidence Rules agenda book" ], "notes": "August 2026 preliminary draft was checked and did not include AI-relevant Rule 707 text.", "url": "https://safelegalai.com/courts/evidence/rules/us-fre-707-proposed", "notice": "Provided as is, without warranty. Not legal advice. SafeLegalAI (Cognesio LLP) records what courts, regulators, legislatures and vendors' own public pages state; the linked official documents are the record. Names and marks belong to their owners. Anyone named may reply: https://safelegalai.com/report. Full terms: https://safelegalai.com/disclaimer", "page": "https://safelegalai.com/courts/evidence/rules/us-fre-707-proposed" }, { "instrument_id": "us-fjc-deepfake-survey-2026", "title": "Federal Judicial Center deepfake survey", "body": "Survey of federal judges about deepfake evidence", "country": "US", "jurisdiction": "us-federal", "state": null, "type": "survey", "status": "published", "status_date": "2026-03-25", "status_quote": "15 of the responding judges have encountered deepfake issues.", "effective_date": null, "next_event": null, "operative_text": "The FJC survey summary reported in the May 2026 Evidence Rules agenda book found that some federal judges had encountered deepfake issues and most respondents favored a rule. The survey informed the committee’s continued work on draft Rule 901(c).", "text_licence": "public-domain", "summary": "The Federal Judicial Center survey was requested to test whether deepfakes are a real courtroom problem. Its summary in the agenda book reports judicial encounters with deepfake issues and majority support for a rule, providing empirical support for further Rule 901(c) consideration.", "source_url": "https://www.uscourts.gov/sites/default/files/document/2026-05-evidence-rules-agenda-book.pdf", "archive_url": "https://web.archive.org/web/20260907222159/https://www.uscourts.gov/sites/default/files/document/2026-05-evidence-rules-agenda-book.pdf", "fetched_at": "2026-09-07T00:00:00Z", "verification": "fetched-and-read", "lead_source": [ "May 2026 Evidence Rules agenda book" ], "notes": null, "url": "https://safelegalai.com/courts/evidence/rules/us-fjc-deepfake-survey-2026", "notice": "Provided as is, without warranty. Not legal advice. SafeLegalAI (Cognesio LLP) records what courts, regulators, legislatures and vendors' own public pages state; the linked official documents are the record. Names and marks belong to their owners. Anyone named may reply: https://safelegalai.com/report. Full terms: https://safelegalai.com/disclaimer", "page": "https://safelegalai.com/courts/evidence/rules/us-fjc-deepfake-survey-2026" }, { "instrument_id": "us-ncsc-unack-ai-evidence-bench-card", "title": "NCSC/TRI bench card: unacknowledged AI-generated evidence", "body": "Bench card for alleged deepfake evidence", "country": "US", "jurisdiction": "us-state", "state": null, "type": "bench-card", "status": "published", "status_date": "2026-01", "status_quote": "Treat the evidence with heightened scrutiny if the source is questionable.", "effective_date": null, "next_event": null, "operative_text": "Courts should ask about source, access, preservation, chain of custody, alterations, corroboration, forensic verification, metadata, and expert testimony.", "text_licence": "quoted-short", "summary": "The NCSC/TRI deepfake bench card supplies questions for evidence offered as authentic but challenged as AI-generated or altered. It directs courts to probe source, custody, preservation, metadata, corroboration, forensic tools, and expert assistance before admitting disputed digital evidence.", "source_url": "https://ncsc.contentdm.oclc.org/digital/api/collection/tech/id/1244/download", "archive_url": "http://web.archive.org/web/20251009075314/https://ncsc.contentdm.oclc.org/digital/api/collection/tech/id/1244/download", "fetched_at": "2026-09-07T00:00:00Z", "verification": "fetched-and-read", "lead_source": [ "NCSC/TRI AI Policy Consortium" ], "notes": "NCSC content treated as copyrighted; only short quotations used.", "url": "https://safelegalai.com/courts/evidence/rules/us-ncsc-unack-ai-evidence-bench-card", "notice": "Provided as is, without warranty. Not legal advice. SafeLegalAI (Cognesio LLP) records what courts, regulators, legislatures and vendors' own public pages state; the linked official documents are the record. Names and marks belong to their owners. Anyone named may reply: https://safelegalai.com/report. Full terms: https://safelegalai.com/disclaimer", "page": "https://safelegalai.com/courts/evidence/rules/us-ncsc-unack-ai-evidence-bench-card" }, { "instrument_id": "us-ncsc-ack-ai-evidence-bench-card", "title": "NCSC/TRI bench card: acknowledged AI-generated evidence", "body": "Bench card for disclosed AI exhibits", "country": "US", "jurisdiction": "us-state", "state": null, "type": "bench-card", "status": "published", "status_date": "2026-01", "status_quote": "ANY USE OF AI-GENERATED EXHIBITS should be clearly labelled as such.", "effective_date": null, "next_event": null, "operative_text": "AI-generated exhibits should be clearly labelled for the fact-finder at all times.", "text_licence": "quoted-short", "summary": "The NCSC/TRI bench card guides judges when AI use is acknowledged, including enhanced images, simulations, data analysis, and biometric tools. It emphasizes labeling, process reliability, preservation of source material, chain of custody, and expert assistance where appropriate.", "source_url": "https://ncsc.contentdm.oclc.org/digital/api/collection/tech/id/1243/download", "archive_url": "https://web.archive.org/web/20260907222229/https://ncsc.contentdm.oclc.org/digital/api/collection/tech/id/1243/download", "fetched_at": "2026-09-07T00:00:00Z", "verification": "fetched-and-read", "lead_source": [ "NCSC/TRI AI Policy Consortium" ], "notes": "NCSC content treated as copyrighted; only short quotations used.", "url": "https://safelegalai.com/courts/evidence/rules/us-ncsc-ack-ai-evidence-bench-card", "notice": "Provided as is, without warranty. Not legal advice. SafeLegalAI (Cognesio LLP) records what courts, regulators, legislatures and vendors' own public pages state; the linked official documents are the record. Names and marks belong to their owners. Anyone named may reply: https://safelegalai.com/report. Full terms: https://safelegalai.com/disclaimer", "page": "https://safelegalai.com/courts/evidence/rules/us-ncsc-ack-ai-evidence-bench-card" }, { "instrument_id": "uk-judiciary-ai-guidance-2025-deepfakes", "title": "UK Judiciary AI Guidance (October 2025), deepfakes", "body": "Judicial guidance on fake AI material", "country": "GB", "jurisdiction": "uk", "state": null, "type": "guidance", "status": "published", "status_date": "2025-10", "status_quote": "Judges should be aware of this new possibility and potential challenges posed by deepfake technology.", "effective_date": null, "next_event": null, "operative_text": "AI tools are now being used to produce fake material, including text, images and video. Courts and tribunals have always had to handle forgeries, and allegations of forgery, involving varying levels of sophistication. Judges should be aware of this new possibility and potential challenges posed by deepfake technology.", "text_licence": "ogl", "summary": "The October 2025 UK judicial guidance warns that AI can create fake text, image, and video material and that judges should be alert to deepfakes and hidden white-text prompts. The judiciary.uk copyright page confirms Crown copyright material is licensed under the Open Government Licence unless otherwise indicated.", "source_url": "https://www.judiciary.uk/guidance-and-resources/artificial-intelligence-ai-judicial-guidance-october-2025/", "archive_url": "http://web.archive.org/web/20260805091942/https://www.judiciary.uk/guidance-and-resources/artificial-intelligence-ai-judicial-guidance-october-2025/", "fetched_at": "2026-09-07T00:00:00Z", "verification": "fetched-and-read", "lead_source": [ "UK Judiciary guidance page", "judiciary.uk copyright page" ], "notes": null, "url": "https://safelegalai.com/courts/evidence/rules/uk-judiciary-ai-guidance-2025-deepfakes", "notice": "Provided as is, without warranty. Not legal advice. SafeLegalAI (Cognesio LLP) records what courts, regulators, legislatures and vendors' own public pages state; the linked official documents are the record. Names and marks belong to their owners. Anyone named may reply: https://safelegalai.com/report. Full terms: https://safelegalai.com/disclaimer", "page": "https://safelegalai.com/courts/evidence/rules/uk-judiciary-ai-guidance-2025-deepfakes" }, { "instrument_id": "us-ca-judicial-council-ai-rule-2025", "title": "California Judicial Council Rule 10.430 and Standard 10.80", "body": "Generative AI guardrails for California court-related work", "country": "US", "jurisdiction": "us-state", "state": "CA", "type": "guidance", "status": "in-force", "status_date": "2025-09-01", "status_quote": "The Artificial Intelligence Task Force recommends adopting one rule of court and one standard of judicial administration.", "effective_date": "2025-09-01", "next_event": null, "operative_text": "Adopt California Rules of Court, rule 10.430 and California Standards of Judicial Administration, standard 10.80, to address the use of generative artificial intelligence for court-related work.", "text_licence": "quoted-short", "summary": "The California Judicial Council adopted Rule 10.430 and Standard 10.80 for responsible generative AI use in court-related work. The framework is not an evidence-admissibility rule, but it addresses oversight, confidentiality, public trust, and disclosure of court-generated AI materials.", "source_url": "https://jcc.legistar.com/LegislationDetail.aspx?ID=7425777&GUID=1091093C-1D8C-42BC-A848-C66A662F53B1", "archive_url": "https://web.archive.org/web/20260907222434/https://jcc.legistar.com/LegislationDetail.aspx?ID=7425777&GUID=1091093C-1D8C-42BC-A848-C66A662F53B1", "fetched_at": "2026-09-07T00:00:00Z", "verification": "fetched-and-read", "lead_source": [ "California Judicial Council AI Task Force page", "User brief" ], "notes": "Official Legistar item was readable. No separate evidence-specific exhibit rule was located; this row records the statewide AI court-work framework requested by the brief.", "url": "https://safelegalai.com/courts/evidence/rules/us-ca-judicial-council-ai-rule-2025", "notice": "Provided as is, without warranty. Not legal advice. SafeLegalAI (Cognesio LLP) records what courts, regulators, legislatures and vendors' own public pages state; the linked official documents are the record. Names and marks belong to their owners. Anyone named may reply: https://safelegalai.com/report. Full terms: https://safelegalai.com/disclaimer", "page": "https://safelegalai.com/courts/evidence/rules/us-ca-judicial-council-ai-rule-2025" }, { "instrument_id": "us-la-act-250-2025", "title": "Louisiana Act 250 / House Bill 178 (2025)", "body": "Louisiana AI/falsified evidence civil procedure amendments", "country": "US", "jurisdiction": "us-state", "state": "LA", "type": "statute", "status": "in-force", "status_date": "2025-08-01", "status_quote": "The Article's use of artificial intelligence is broad and encompasses the suspected use of deepfakes.", "effective_date": "2025-08-01", "next_event": null, "operative_text": "An attorney shall exercise reasonable diligence to verify the authenticity of evidence before offering it to the court. If an attorney knew or should have known through the exercise of reasonable diligence that evidence was false or artificially manipulated, the offering of that evidence without disclosure of that fact shall be considered a violation of this Article.", "text_licence": "public-domain", "summary": "Louisiana Act 250 amends civil procedure to require attorney diligence before offering evidence and pretrial disclosure or challenge of exhibits suspected to be falsified, AI-generated, or altered. It adds explicit deepfake-related comments to pretrial authenticity and admissibility management.", "source_url": "https://legis.la.gov/Legis/ViewDocument.aspx?d=1425558", "archive_url": "https://web.archive.org/web/20260817152331/https://www.legis.la.gov/legis/ViewDocument.aspx?d=1425558", "fetched_at": "2026-09-07T00:00:00Z", "verification": "fetched-and-read", "lead_source": [ "User brief", "Louisiana Legislature/WAYBACK PDF" ], "notes": "Direct legis.la.gov failed DNS from this environment; a Wayback PDF copy was fetched and read.", "url": "https://safelegalai.com/courts/evidence/rules/us-la-act-250-2025", "notice": "Provided as is, without warranty. Not legal advice. SafeLegalAI (Cognesio LLP) records what courts, regulators, legislatures and vendors' own public pages state; the linked official documents are the record. Names and marks belong to their owners. Anyone named may reply: https://safelegalai.com/report. Full terms: https://safelegalai.com/disclaimer", "page": "https://safelegalai.com/courts/evidence/rules/us-la-act-250-2025" }, { "instrument_id": "in-bsa-2023-section-63", "title": "Bharatiya Sakshya Adhiniyam, 2023 section 63", "body": "Electronic-record admissibility rule", "country": "IN", "jurisdiction": "india", "state": null, "type": "rule", "status": "in-force", "status_date": "2024-07-01", "status_quote": null, "effective_date": "2024-07-01", "next_event": null, "operative_text": "Electronic or digital records are admissible as documents if the statutory computer-output conditions and certificate requirements in section 63 are satisfied.", "text_licence": "quoted-short", "summary": "Section 63 is not AI-specific, but it governs admissibility of electronic and digital records in India through computer-output conditions and certificates. It is included as a baseline electronic-records rule relevant to AI-generated or AI-processed evidence.", "source_url": "https://www.indiacode.nic.in/handle/123456789/20063?sam_handle=123456789/1362", "archive_url": null, "fetched_at": "2026-09-07T00:00:00Z", "verification": "link-only", "lead_source": [ "India Code official portal", "User brief" ], "notes": "India Code returned HTTP 403 to fetch; section text should be verified manually from the official portal or Gazette.", "url": "https://safelegalai.com/courts/evidence/rules/in-bsa-2023-section-63", "notice": "Provided as is, without warranty. Not legal advice. SafeLegalAI (Cognesio LLP) records what courts, regulators, legislatures and vendors' own public pages state; the linked official documents are the record. Names and marks belong to their owners. Anyone named may reply: https://safelegalai.com/report. Full terms: https://safelegalai.com/disclaimer", "page": "https://safelegalai.com/courts/evidence/rules/in-bsa-2023-section-63" } ] }