{"meta":{"exportedAt":"2026-11-07T02:55:54.741Z","formatVersion":2,"selection":{"q":"privacy","system":"","harm":"","context":"","country":"","role":"","relation":"contextual","evidence":"","year":"2024","response":"","severity":"","verification":"","view":"incidents","sort":"added"},"totalIncidents":8,"coverage":{"cases":8,"countries":4,"languages":4,"unknownLocation":2,"locationPending":0,"unknownLanguage":0,"unknownDate":0,"lawsuits":4,"regulatory":2,"minors":1,"coreRelations":1,"contextualRelations":8,"mixedRelations":1,"unknownRelations":0,"relationPending":0,"relationUnknown":0},"countingNote":"Distinct public cases in this selection. People counts apply within individual cases only; cross-case person overlap has not been resolved. No population incidence estimate.","affectedCountNote":"Interpret person counts with affectedCountStatus and the reported effects. Unquantified zeros are placeholders, not a measured zero.","source":"AI incidents","publisher":"NOPE","url":"https://nope.net/incidents","license":"CC BY 4.0"},"incidents":[{"id":"2024-gemini-in-google-drive-summarised-users-opened-tax-return-unrequested-with-summary-settings-off-first-person","caseFacts":{"claims":[{"id":"c1","status":"reported","evidence":[{"locator":"'Just pulled up my tax return in @Google Docs--and unbidden, Gemini summarized it'; 'appears this is happening with any PDF of mine that I open from Drive'","relation":"supports","source_id":"s1"},{"locator":"'Just pulled up my tax return in @Google Docs--and unbidden, Gemini summarized it'; 'this is supposedly happening within the larger umbrella of Google Drive and not Google Docs specifically'","relation":"supports","source_id":"s2"}],"assertion":"The account holder says that on 10 July 2024 he opened his tax return and Gemini summarised it without his asking; his later testing in the thread places this in the PDF viewer of Google Drive.","causal_attribution":"Account holder's own report of the AI output; the summary itself was not inspected."},{"id":"c2","status":"reported","evidence":[{"locator":"'and it was already turned off! Yet Gemini is absolutely giving me document summaries'","relation":"supports","source_id":"s1"},{"locator":"'only to find that Gemini summaries in Gmail, Drive, and Docs were already disabled'","relation":"supports","source_id":"s2"},{"locator":"'found the control, which was already disabled'","relation":"supports","source_id":"s3"}],"assertion":"He says the setting meant to keep Gemini out of his documents was already turned off while Gemini kept giving him document summaries.","causal_attribution":"Account holder's report; why the setting did not apply is not established."},{"id":"c3","status":"reported","evidence":[{"locator":"'So I asked Gemini to point me to the settings. This is what it said. Tried both, neither works'","relation":"supports","source_id":"s1"},{"locator":"'AI chatbot on two occasions to pinpoint the settings'","relation":"supports","source_id":"s3"}],"assertion":"He says he asked Gemini where the setting was and neither of the two locations it gave worked.","causal_attribution":"Account holder's report of Gemini's replies."},{"id":"c4","status":"reported","evidence":[{"locator":"'If you ever used Gemini with a particular PDF, it will re-invoke for every other PDF you load until you close it again'; 'My working theory until Google can clarify is that using Workplace Labs overrode the explicit setting turning off Workplace integrations of Gemini'","relation":"supports","source_id":"s1"},{"locator":"'He additionally theorizes that it may have been caused by him enabling Google Workspace Labs back in 2023'","relation":"supports","source_id":"s2"},{"locator":"'ativou essas funcionalidades em 2023, antes de elas serem liberadas para todos os usu'","relation":"supports","source_id":"s4"}],"assertion":"In later posts he says that once Gemini was used with one PDF it was invoked again for every PDF he opened until closed, and puts forward a working theory that his 2023 Workspace Labs sign-up overrode the opt-out.","causal_attribution":"The account holder's own explanation, offered as a theory; it bears on whether he had enabled the feature."},{"id":"c5","status":"reported","evidence":[{"locator":"'I just went to the chat logs to delete the doc that was uploaded and the summary given in response'","relation":"supports","source_id":"s1"}],"assertion":"He says that when he went to his Gemini activity to delete the uploaded document and the summary, they were not shown there.","causal_attribution":"Account holder's report; what Google retained is not established."},{"id":"c6","status":"reported","evidence":[{"locator":"'Using Gemini in Google Workspace requires a user to proactively enable it'; 'may have used Gemini via the side panel in Drive and that closing the panel might solve the issue'","relation":"supports","source_id":"s2"},{"locator":"'Using Gemini in Google Workspace requires a user to proactively enable it'","relation":"supports","source_id":"s3"}],"assertion":"A Google spokesperson said that using Gemini in Google Workspace requires a user to enable it proactively and that content is not otherwise stored without permission; Tom's Hardware reports Google suggested he may have used Gemini through the Drive side panel and that closing it might solve the issue.","causal_attribution":"Google's statement disputes that the processing was unconsented."}],"effects":[{"label":"private tax return summarised by Gemini without the account holder's request while his opt-out setting was off (his account)","claim_id":"c2","direction":"negative"}],"sources":[{"id":"s1","url":"https://x.com/KevinBankston/status/1811075836558114968","kind":"social_media_post","access":"read","language":"en","translation_note":"Read in English by the research agent (an AI) on 2026-11-01: the first post through the fxtwitter API (HTTP 200) and the whole 36-post thread through the Thread Reader App unroll (threadreaderapp.com, HTTP 200); no translation and no human reviewer. Attached screenshots were not inspected.","independence_group":"technology-adviser-x-thread-gemini-drive-summary"},{"id":"s2","url":"https://www.tomshardware.com/tech-industry/artificial-intelligence/gemini-ai-caught-scanning-google-drive-hosted-pdf-files-without-permission-user-complains-feature-cant-be-disabled","kind":"news_report","access":"read","language":"en","translation_note":"Read in English by the research agent (an AI) on 2026-11-01 (HTTP 200); no translation. Relays the thread and carries Google's statement and suggestion to Tom's Hardware.","independence_group":"technology-adviser-x-thread-gemini-drive-summary"},{"id":"s3","url":"https://www.techradar.com/pro/security/gemini-ai-platform-caught-scanning-google-drive-files-without-user-permission","kind":"news_report","access":"read","language":"en","translation_note":"Read in English by the research agent (an AI) on 2026-11-01 (HTTP 200); no translation. Relays the thread and carries the same Google statement.","independence_group":"technology-adviser-x-thread-gemini-drive-summary"},{"id":"s4","url":"https://tecnoblog.net/noticias/gemini-le-documento-salvo-no-google-drive-contra-a-vontade-de-usuario/","kind":"news_report","access":"read","language":"pt","translation_note":"Read in Portuguese by the research agent (an AI) on 2026-11-01 (HTTP 200); passages translated by the research agent, no human translator. Relays the thread and carries the same Google statement sent to Google Brasil's press office.","independence_group":"technology-adviser-x-thread-gemini-drive-summary"}],"version":1,"ai_roles":["own_use"],"contexts":["privacy","finance","everyday_life"],"unknowns":["Whether the account holder had enabled Gemini for Drive at some point, through Workspace Labs or an earlier use of the side panel.","Why the opt-out setting did not stop the summaries, and whether Google changed anything after the report.","Whether Google retained the document content or the summary."],"geography":{"basis":"No inspected source states where the account holder was when he opened the document; his employer's location does not locate the event. No country is inferred.","court_countries":[],"event_countries":[],"affected_person_countries":[]},"publication":{"basis":"Published under the 2026-10-06 privacy clarification and the public-forum rule as a first-person account of a private financial document processed by an AI assistant without the person's request and against an opt-out setting. Claims are attributed to the account holder; Google's statement and his own alternative explanations are recorded.","reviewed_on":"2026-11-01"},"ai_involvement":{"basis":"The account holder says Gemini generated a summary of his tax return when he opened it, without his request, while the setting meant to keep Gemini out of his documents was off; the unrequested processing of his private financial document by the AI assistant is the reported privacy harm. Google says Gemini in Workspace must be enabled by the user, and the account holder himself suggests an earlier Workspace Labs sign-up or an earlier use of the side panel may explain it, so whether he had enabled the feature is disputed.","status":"reported"},"person_relations":["received_personal_information"]},"name":"AI governance adviser says Gemini in Google Drive summarised his tax return unrequested while its summary setting was off; Google says users must enable Gemini","summary":"In a thread on X on 10 July 2024, an AI governance adviser wrote that he had just opened his tax return and that Gemini, Google's AI assistant, summarised it without being asked. He wrote that he could not find a setting to stop it, that the two locations Gemini itself suggested did not work, and that when another user showed him the setting meant to keep Gemini out of his documents it was already turned off while Gemini kept giving document summaries. His later testing, in the same thread, suggested that once Gemini had been opened for one PDF in Google Drive it was invoked again for every PDF he opened until he closed it, and he put forward a working theory that his 2023 sign-up for Workspace Labs had overridden the opt-out. A Google spokesperson told technology outlets that using Gemini in Google Workspace requires a user to enable it proactively and that content is not stored without permission; Tom's Hardware reports that Google suggested he may have used Gemini through the Drive side panel and that closing the panel might solve the issue.","incidentDate":"2024-07-10","incidentKind":"single_event","incidentDatePrecision":"day","exposurePattern":"single_interaction","reportedDate":"2024-07-10","aiSystem":"Gemini assistant in Google Workspace (Google Drive PDF viewer side panel and Google Docs), as described by the account holder","aiProduct":"Gemini (reported)","severity":"low","verificationStatus":"unverified","harmCategories":[],"harmOutcomes":["other_material_harm"],"harmOutcomeSummary":"The account holder says Gemini summarised his private tax return when he opened it, without his request and while the setting meant to keep Gemini out of his documents was off (first-person account; Google says Gemini in Workspace must be enabled by the user).","frameworkFacets":[],"causationStatus":"alleged","participantUsersAffectedMin":1,"otherPeopleHarmedMin":0,"affectedCountStatus":"exact","affectedCountEvidence":"One person: the account holder whose own tax return was summarised. His question about how many other people were affected is a question, not a count.","victimAgeRange":"unknown","platformType":"other","outcomeType":"media_coverage","outcomeStatus":"unknown","primarySourceUrl":"https://x.com/KevinBankston/status/1811075836558114968","primarySourceLabel":"Thread on X by the account holder, 10 Jul 2024: Gemini summarised his tax return unbidden; opt-out setting found already off","firstPublishedAt":"2026-11-01T03:07:17.532166+00:00","updatedAt":"2026-11-01T03:07:17.532166+00:00","scopeVersion":"facts-v3","tags":["first-person","x-twitter","gemini","google-drive","privacy","consent","opt-out-setting"]},{"id":"2024-lawrence-kansas-gaggle-ai-flagged-students-class-photos-as-child-sexual-images-pupils-questioned-photos-removed","caseFacts":{"claims":[{"id":"c1","status":"reported","evidence":[{"locator":"'In November 2023, USD 497 and its Board of Education implemented the use of Gaggle'; 'Gaggle operates through artificial intelligence that flags items for review. Gaggle moderators conduct a brief contextual review to determine the propriety of the content at issue and quarantine the item unless the flag is deemed a false positive.'","relation":"supports","source_id":"s1"},{"locator":"'challenges the district’s 2023 decision to use a third-party artificial intelligence-powered software service called Gaggle to monitor student usage of district-issued computers'","relation":"supports","source_id":"s4"}],"assertion":"The Lawrence school district began using Gaggle to monitor student school accounts in November 2023; the plaintiffs' complaint says Gaggle operates through artificial intelligence that flags items for review by Gaggle moderators.","causal_attribution":"The complaint's description of the system; the newspaper describes the service the same way. Gaggle's own account of how the flag arose is not in the inspected sources."},{"id":"c2","status":"reported","evidence":[{"locator":"'In early 2024, District administrators summoned art, journalism, and photography students to the front office at Lawrence High School for questioning.'; 'administrators told the group that their images had been flagged by Gaggle for indecent exposure and child pornography.'; 'admitted they had not seen the content at issue, but rather, that they were informed second-hand by Gaggle that the content had been flagged.'","relation":"supports","source_id":"s1"},{"locator":"'told the group that Gaggle had flagged their images for indecent exposure and child pornography.'","relation":"supports","source_id":"s2"}],"assertion":"According to the complaint, in early 2024 administrators summoned art, journalism and photography students to the front office and told them their images had been flagged by Gaggle for indecent exposure and child pornography, while admitting they had not seen the flagged content themselves.","causal_attribution":"Plaintiffs' allegations; the court's order summarises the amended complaint and does not find these facts."},{"id":"c3","status":"reported","evidence":[{"locator":"'of having photos or documents on her computer that constituted indecent exposure or child pornography.'; 'she had uploaded a photography assignment that included a picture of one of her friends who was fully clothed and wearing a spaghetti-strap tank top.'; 'to recall and explain from memory minute details regarding the photographs she had taken, but she was not allowed to view or display those photos.'; 'her detention and interrogation by Assistant Principal'","relation":"supports","source_id":"s1"}],"assertion":"According to the complaint, an assistant principal accused a photography student of having images constituting indecent exposure or child pornography, refused to let her view the images, and questioned her about them from memory; she told him she had uploaded that day a photography assignment showing a fully clothed friend.","causal_attribution":"A plaintiff's account in the complaint."},{"id":"c4","status":"reported","evidence":[{"locator":"'determined that the flagged content consisted of original artistic photographs taken and uploaded for class assignments.'; 'discovered that multiple digital photographs had been removed from their school accounts.'; 'would trigger additional scrutiny, interrogations, deletions, and/or discipline.'; 'and other students later confirmed the seizures by comparing their school accounts to independent backup drives.'; 'was subsequently required to meet with District staff and explain each image in her photography portfolio.'; 'Only after months of student advocacy did the District restore the removed images.'","relation":"supports","source_id":"s1"}],"assertion":"According to the complaint, the flagged content consisted of original photographs taken for class assignments; the photographs were removed from the students' school accounts without notice, the students found this by comparing their accounts with personal backups, one student had to explain each image in her portfolio to district staff and became concerned that future assignments would trigger more scrutiny, and the district restored the images only after months of advocacy.","causal_attribution":"Plaintiffs' allegations; the removal is also reported by AP (c5)."},{"id":"c5","status":"reported","evidence":[{"locator":"'Students in one photography class were called to the principal's office over concerns Gaggle had detected nudity. The photos had been automatically deleted from the students' Google Drives, but students who had backups of the flagged images on their own devices showed it was a false alarm. District officials said they later adjusted the software's settings to reduce false alerts.'","relation":"supports","source_id":"s3"},{"locator":"'Students in one photography class were called to the principal’s office over concerns Gaggle had detected nudity.'","relation":"supports","source_id":"s4"}],"assertion":"The Associated Press reported that students in a photography class were called to the principal's office over concerns Gaggle had detected nudity, that the photos had been automatically deleted from their Google Drives, that students with backups showed it was a false alarm, and that district officials said they later adjusted the software's settings to reduce false alerts.","causal_attribution":"AP's reporting, relayed by the Journal-World from AP's review; the settings change is the district officials' statement to AP. Both outlets carry one AP origin."},{"id":"c6","status":"documented","evidence":[{"locator":"'Plaintiffs have not cited (and the Court has not found) any law that when AI surveillance flags student material as potentially criminal, a school official cannot reasonably rely on that information to call in a student for questioning.'; 'the claims for injunctive and declaratory relief of the graduated students, the claims against the School Board and, based on qualified immunity, all claims against'","relation":"supports","source_id":"s2"}],"assertion":"On 8 April 2026 the court dismissed, on qualified immunity, all claims against the assistant principal and another official, saying it had found no law that when AI surveillance flags student material as potentially criminal a school official cannot reasonably rely on it to call in a student for questioning; it also dismissed the graduated students' claims for injunctive and declaratory relief and the claims against the school board.","causal_attribution":"The court's ruling; it decides qualified immunity on the pleadings and makes no finding on whether the flags were accurate."},{"id":"c7","status":"documented","evidence":[{"locator":"'Date of Last Known Filing: Oct. 26, 2026'; 'Notice to Take Deposition'","relation":"supports","source_id":"s5"}],"assertion":"The case remained open in late October 2026, with depositions noticed and a last known filing on 26 October 2026.","causal_attribution":"Docket record."}],"effects":[{"label":"students questioned over an AI flag calling their class photography indecent exposure or child pornography, with their photographs removed for months","claim_id":"c4","direction":"negative"}],"sources":[{"id":"s1","url":"https://storage.courtlistener.com/recap/gov.uscourts.ksd.159249/gov.uscourts.ksd.159249.1.0.pdf","kind":"court_filing","access":"read","language":"en","translation_note":"Read in English on 2026-10-27; no translation.","independence_group":"dkan-2-25-cv-02428-complaint"},{"id":"s2","url":"https://storage.courtlistener.com/recap/gov.uscourts.ksd.159249/gov.uscourts.ksd.159249.56.0.pdf","kind":"court_order","access":"read","language":"en","translation_note":"Read in English on 2026-10-27; no translation.","independence_group":"dkan-2-25-cv-02428-order"},{"id":"s3","url":"https://www.yahoo.com/news/articles/schools-using-ai-surveillance-protect-110342650.html","kind":"news_report","access":"read","language":"en","translation_note":"Read in English on 2026-10-27; no translation.","independence_group":"ap-gaggle-2025"},{"id":"s4","url":"https://www2.ljworld.com/news/schools/2025/aug/07/lawrence-school-district-sued-in-federal-court-for-use-of-ai-powered-surveillance-system-students-claim-gaggle-results-in-illegal-searches/","kind":"news_report","access":"read","language":"en","translation_note":"Read in English on 2026-10-27; no translation.","independence_group":"ljworld-2025-08-07"},{"id":"s5","url":"https://www.courtlistener.com/docket/70989873/","kind":"court_docket","access":"read","language":"en","translation_note":"Read in English on 2026-10-27; no translation.","independence_group":"courtlistener-docket"}],"version":1,"ai_roles":["institutional_use"],"contexts":["education","public_services","privacy","justice"],"unknowns":["Exactly when in early 2024 the questioning took place and how many students were summoned.","Whether Gaggle's human moderators reviewed the specific photographs before the district was notified, and what they concluded.","Whether the photographs were removed automatically by Gaggle or by a reviewer; AP says automatically deleted, the complaint describes moderator quarantine.","Gaggle's and the district's responses to these specific allegations in the litigation.","The outcome of the remaining claims against the district."],"geography":{"basis":"The complaint places the questioning at a Lawrence, Kansas, high school of the district's own school accounts; the suit is in the federal District of Kansas.","court_countries":["US"],"event_countries":["US"],"affected_person_countries":["US"]},"publication":{"basis":"Published as a school's use of an AI monitoring flag about students: the complaint and AP describe Gaggle flagging photography students' class images as nudity or child pornography, the students' questioning and the removal of their work, district officials told AP it was a false alarm, and a federal court has ruled on the official's reliance on the AI flag.","reviewed_on":"2026-10-27"},"ai_involvement":{"basis":"The complaint says Gaggle operates through artificial intelligence that flags items for review and that administrators told the students their images had been flagged by Gaggle for indecent exposure and child pornography, then questioned them without having seen the images; AP reports Gaggle detected nudity in the photography class's uploads, that district officials called it a false alarm and adjusted the software's settings, and the court's order treats the flag as an AI surveillance flag on which the official relied. Gaggle moderators review flags before schools are notified, and no inspected source says what each reviewer saw.","status":"reported"},"person_relations":["made_claim_about","informed_decision_about"]},"name":"Lawrence, Kansas: students say Gaggle AI monitoring flagged class photography as child sexual images, leading to questioning and removed work","summary":"The Lawrence, Kansas, school district began using Gaggle, a student-monitoring service, on school accounts in November 2023. A federal lawsuit filed on 1 August 2025 by current and former students alleges that Gaggle, which the complaint says operates through artificial intelligence that flags items for review by Gaggle moderators, flagged original photography assignments uploaded by art and photography students. According to the complaint, in early 2024 security officers took students out of class to the front office of a Lawrence high school, where administrators told them their images had been flagged by Gaggle for indecent exposure and child pornography; an assistant principal accused a photography student of having such material, refused to let her view the images, and questioned her about them from memory, and administrators admitted they had not seen the flagged content. The student told them she had uploaded that day a photography assignment showing a fully clothed friend; the complaint says the flagged content consisted of original photographs taken for class assignments, that students found their photographs removed from their school accounts only by comparing them with personal backups, that one student had to explain each image in her portfolio to district staff and became concerned that future assignments would bring more scrutiny, and that the district restored the images only after months of advocacy. The Associated Press, in reporting that included an analysis of the district's Gaggle alert records, said the photos had been automatically deleted and that students with backups showed it was a false alarm, and that district officials said they later adjusted the software's settings to reduce false alerts. On 8 April 2026 the federal court dismissed, on qualified immunity, the claims against the assistant principal and another official, saying it had found no law establishing that a school official cannot rely on an AI surveillance flag to call in a student for questioning; other claims against the district continue, and the docket shows depositions noticed in October 2026.","incidentDate":"2024-01-01","incidentKind":"single_event","incidentDatePrecision":"year","exposurePattern":"single_interaction","reportedDate":"2025-08-01","aiSystem":"Gaggle Safety Management, a student-monitoring service that scans school email and files; the complaint says it operates through artificial intelligence that flags items for review by Gaggle moderators, and the district's board agenda described trained safety experts and an algorithm","aiProduct":"Gaggle","aiCompany":"Gaggle.Net, Inc.","severity":"medium","verificationStatus":"credible","harmCategories":[],"harmOutcomes":["educational_harm","psychological_distress"],"harmOutcomeSummary":"According to the plaintiffs' complaint, students were taken out of class and questioned after Gaggle flagged their photography assignments as indecent exposure or child pornography, their photographs were removed from their school accounts without notice and restored only after months, one student had to explain every image in her portfolio, and she became concerned that future assignments would bring more scrutiny. District officials told the Associated Press it was a false alarm and that settings were adjusted.","frameworkFacets":[],"causationStatus":"alleged","participantUsersAffectedMin":0,"otherPeopleHarmedMin":2,"affectedCountStatus":"partial","affectedCountEvidence":"Two photography students who are plaintiffs describe being questioned and having photographs removed (complaint paras 93 to 126). The complaint and AP describe other art, journalism and photography students summoned or flagged without a count, so they are not quantified. The 1,200 alerts in AP's records analysis are alerts, not harmed people, and are not counted.","victimAgeRange":"unknown","jurisdiction":"US","platformType":"other","outcomeType":"lawsuit_ongoing","outcomeStatus":"ongoing","primarySourceUrl":"https://storage.courtlistener.com/recap/gov.uscourts.ksd.159249/gov.uscourts.ksd.159249.1.0.pdf","primarySourceLabel":"Students' complaint against the Lawrence school district over Gaggle, D. Kan. 2:25-cv-02428, filed 1 August 2025","firstPublishedAt":"2026-10-27T03:08:48.076061+00:00","updatedAt":"2026-10-27T03:08:48.076061+00:00","scopeVersion":"facts-v3","tags":["gaggle","school-monitoring","student-surveillance","false-flag","photography","qualified-immunity","kansas"]},{"id":"2024-london-bridge-met-police-live-facial-recognition-wrongly-matched-man-to-watchlisted-brother-stop-settlement","caseFacts":{"claims":[{"id":"c1","status":"documented","evidence":[{"locator":"'It is not in dispute that, on 23 February 2024, Mr Thompson passed an LFR deployment in the area around London Bridge. His face was matched to an image contained on the watchlist in use on that day, so as to generate an alert. The LFR system had detected a high level of facial similarity between images obtained of Mr Thompson and the image of his brother who was on the watchlist as being the subject of an outstanding arrest warrant'","relation":"supports","source_id":"s1"},{"locator":"'Mr Thompson was mistakenly identified during an LFR deployment near London Bridge after being matched to an image of his brother, who was on a police watchlist.'","relation":"supports","source_id":"s2"}],"assertion":"On 23 February 2024 the man passed a Met Police live facial recognition deployment near London Bridge; the system detected a high level of facial similarity between images of him and the watchlist image of his brother, who was the subject of an outstanding arrest warrant, and generated an alert.","causal_attribution":"Facts recorded by the court as not in dispute; the press summary is the court's own summary of the same judgment."},{"id":"c2","status":"reported","evidence":[{"locator":"'Mr Thompson was stopped, detained and questioned by police officers after having been matched by LFR'; 'they nevertheless scrutinised him, asked questions and made him prove his identity. When he declined to allow his fingerprints to be taken, he was threatened with arrest. Mr Thompson explains how the experience caused him to feel angry and distressed.'","relation":"supports","source_id":"s1"},{"locator":"'He was held by officers for almost 30 minutes, who repeatedly demanded scans of his fingerprints and threatened him with arrest'","relation":"supports","source_id":"s3"}],"assertion":"According to the court's account of his case, officers stopped, detained and questioned him, made him prove his identity, and threatened him with arrest when he declined to have his fingerprints taken; he says the experience left him angry and distressed.","causal_attribution":"The court's introduction summarises his account; Big Brother Watch relays the same account. Both rest on his evidence, so the claim is reported, and the 30-minute figure is attributed to Big Brother Watch."},{"id":"c3","status":"documented","evidence":[{"locator":"'They ceased their engagement with him after considering information held on the Police National Computer about his brother. Mr Thompson was able to show police a copy of his passport photograph on his phone, which enabled his identity to be confirmed.'","relation":"supports","source_id":"s1"}],"assertion":"Officers ended their engagement after considering police records about his brother, and he confirmed his identity with a copy of his passport photograph on his phone.","causal_attribution":"Facts set out in the judgment's account of the first claimant."},{"id":"c4","status":"documented","evidence":[{"locator":"'On 18 December 2024 a consent order was agreed (eventually sealed on 30 April 2025) by which grounds 1 and 2 of the original claim were withdrawn on terms as to costs and a settlement sum payable to Mr Thompson.'","relation":"supports","source_id":"s1"}],"assertion":"His original judicial review grounds about the stop were withdrawn under a consent order agreed on 18 December 2024 and sealed on 30 April 2025, on terms as to costs and a settlement sum payable to him.","causal_attribution":"Procedural history recorded in the judgment; the amount and any admission are not stated."},{"id":"c5","status":"documented","evidence":[{"locator":"'Accordingly, the claim is dismissed.'","relation":"supports","source_id":"s1"},{"locator":"'The court emphasised that it is not concerned with the merits of facial recognition technology as a policing tool.'","relation":"supports","source_id":"s2"}],"assertion":"On 21 April 2026 the Divisional Court dismissed the amended claim challenging the Met's revised September 2024 live facial recognition policy, saying it was not concerned with the merits of the technology as a policing tool.","causal_attribution":"Outcome of the judgment; the challenge concerned the revised policy, not the lawfulness of the February 2024 stop."},{"id":"c6","status":"reported","evidence":[{"locator":"'The claimants have announced plans to appeal.'; 'Thompson added: “No one should be treated like a criminal due to a computer error.”'","relation":"supports","source_id":"s4"}],"assertion":"IDTechWire reported on 22 April 2026 that the claimants planned to appeal, quoting him: 'No one should be treated like a criminal due to a computer error.'","causal_attribution":"Trade press report of the claimants' statements; no appeal decision was found."}],"effects":[{"label":"stopped, questioned and threatened with arrest after a live facial recognition false match, and left angry and distressed","claim_id":"c2","direction":"negative"}],"sources":[{"id":"s1","url":"https://caselaw.nationalarchives.gov.uk/ewhc/admin/2026/915","kind":"court_judgment","access":"read","language":"en","translation_note":"Read in English on 2026-10-26; no translation.","independence_group":"ewhc-2026-915"},{"id":"s2","url":"https://www.judiciary.uk/wp-content/uploads/2026/04/Summary-of-judgment-Thomson-and-Carlo-v-Metropolitan-Police-Commmissioner-AC-2024-LON-001764-21-04-2026.pdf","kind":"court_press_summary","access":"read","language":"en","translation_note":"Read in English on 2026-10-26; no translation.","independence_group":"ewhc-2026-915"},{"id":"s3","url":"https://bigbrotherwatch.org.uk/press-releases/met-police-face-major-legal-challenge-over-use-of-live-facial-recognition-technology/","kind":"advocacy_press_release","access":"read","language":"en","translation_note":"Read in English on 2026-10-26; no translation.","independence_group":"big-brother-watch"},{"id":"s4","url":"https://idtechwire.com/uk-high-court-upholds-met-police-lfr-policy-claimants-plan-appeal/","kind":"news_report","access":"read","language":"en","translation_note":"Read in English on 2026-10-26; no translation.","independence_group":"idtechwire-2026-04-22"}],"version":1,"ai_roles":["institutional_use"],"contexts":["justice","privacy","public_services"],"unknowns":["How long the stop lasted according to the police; only Big Brother Watch gives almost 30 minutes.","The settlement amount and whether the force accepted any fault in the consent order.","Whether the claimants' announced appeal was filed or permitted.","Paragraph 20 of the judgment dates the use of LFR complained of to 23 February 2023; paragraphs 17 and 141 give 23 February 2024, which is used here."],"geography":{"basis":"The judgment places the deployment and stop near London Bridge station and says he was born and lives in London; the judicial review was decided by the High Court in London.","court_countries":["GB"],"event_countries":["GB"],"affected_person_countries":["GB"]},"publication":{"basis":"Published as a facial recognition false match leading to a police stop, with a legal outcome: the High Court judgment records the system's alert matching him to his brother's watchlist image as not in dispute, a consent order with a settlement sum paid to him, and the dismissal of his later challenge to the force's deployment policy.","reviewed_on":"2026-10-26"},"ai_involvement":{"basis":"The High Court records as not in dispute that the Met's live facial recognition system detected a high level of facial similarity between images of him and his brother's watchlist image and generated the alert; the court's account and Big Brother Watch describe officers stopping and questioning him after that match. The system's false match is the output that went wrong; the officers' decision to stop him and demand proof of identity followed from it.","status":"supported"},"person_relations":["made_claim_about","informed_decision_about"]},"name":"London: Met Police live facial recognition wrongly matched a man to his watchlisted brother; officers stopped him and he later received a settlement","summary":"On 23 February 2024 Shaun Thompson, a London community volunteer whom the 2026 judgment describes as 39, walked past a Metropolitan Police live facial recognition deployment near London Bridge station. The High Court records as not in dispute that the system detected a high level of facial similarity between images of him and the watchlist image of his brother, who was the subject of an outstanding arrest warrant, and generated an alert. According to the court's account, officers stopped, detained and questioned him and made him prove his identity, and when he declined to have his fingerprints taken he was threatened with arrest; he says the experience left him angry and distressed. Officers ended the engagement after checking police records about his brother, and he confirmed his identity with a passport photograph on his phone. Big Brother Watch, whose director was his co-claimant, says he was held for almost 30 minutes despite showing several identity documents. His judicial review claim about the stop, issued in May 2024, was withdrawn under a consent order agreed on 18 December 2024 on terms including a settlement sum paid to him. An amended challenge to the Met's revised live facial recognition policy of September 2024, brought with Big Brother Watch's director, was dismissed by the Divisional Court on 21 April 2026; the court said it was not judging the merits of the technology, and the claimants announced plans to appeal.","incidentDate":"2024-02-23","incidentKind":"single_event","incidentDatePrecision":"day","exposurePattern":"single_interaction","reportedDate":"2025-07-22","aiSystem":"The Metropolitan Police Service's live facial recognition system, which compares faces captured in public places with biometric data from a watchlist of sought persons; the judgment says the force's operational LFR used technology known as the NEC-3 algorithm","aiProduct":"Metropolitan Police live facial recognition","severity":"medium","verificationStatus":"credible","harmCategories":[],"harmOutcomes":["loss_of_liberty","psychological_distress"],"harmOutcomeSummary":"A live facial recognition alert wrongly matched him to his brother's watchlist image; according to the High Court's account he was stopped, detained and questioned, made to prove his identity and threatened with arrest when he declined fingerprints, and he says he was left angry and distressed. Big Brother Watch says he was held for almost 30 minutes.","frameworkFacets":[],"causationStatus":"established","participantUsersAffectedMin":0,"otherPeopleHarmedMin":1,"affectedCountStatus":"exact","affectedCountEvidence":"One person: the first claimant, whose face the system matched to his brother's watchlist image and whom officers stopped. The co-claimant witnessed the stop and is not counted.","victimAgeRange":"adult","jurisdiction":"GB","platformType":"other","outcomeType":"lawsuit_dismissed","outcomeStatus":"ongoing","primarySourceUrl":"https://caselaw.nationalarchives.gov.uk/ewhc/admin/2026/915","primarySourceLabel":"R (Thompson and Carlo) v Commissioner of Police of the Metropolis [2026] EWHC 915 (Admin), 21 April 2026","firstPublishedAt":"2026-10-26T03:13:03.528278+00:00","updatedAt":"2026-10-26T03:13:03.528278+00:00","scopeVersion":"facts-v3","tags":["live-facial-recognition","police-stop","false-match","watchlist","judicial-review","settlement"]},{"id":"2024-germany-microsoft-copilot-falsely-described-court-reporter-as-criminal-and-disclosed-his-address","caseFacts":{"claims":[{"id":"c1","status":"reported","evidence":[{"locator":"'ein Ausbrecher aus der Psychiatrie und ein Witwenbetrüger. All das waren Angeklagte in Verfahren, über die'; 'als Journalist berichtet hat.'; 'wurde in einem Missbrauchs-Fall gegen Kinder und Schutzbefohlene angeklagt.'","relation":"supports","source_id":"s1"},{"locator":"'eine erschreckende Biografie über seine \"kriminelle Vergangenheit\" erhalten.'; 'Im Rahmen seiner journalistischen Tätigkeit als Gerichtsreporter hat er aber über diese Fälle berichtet.'","relation":"supports","source_id":"s2"},{"locator":"'AI chatbot had blamed him for crimes he had covered.'","relation":"supports","source_id":"s3"}],"assertion":"Asked about a court reporter, Copilot described him as a man charged in a child abuse case, an escapee from psychiatry and a fraudster who cheated widows; all were defendants in proceedings he had reported on.","causal_attribution":"The reporter's account (to SWR, relayed by heise and ComputerBild, and to The Register); one witness, so reported."},{"id":"c2","status":"reported","evidence":[{"locator":"'Der Copilot liefert dann auch noch die volle Adresse des betroffenen Journalisten, samt Telefonnummer und auf Wunsch einer Routenplanung.'","relation":"supports","source_id":"s1"},{"locator":"'verrät auch die private Telefonnummer des Journalisten, die volle Wohnadresse und bietet obendrein eine Routenplanung dorthin an.'","relation":"supports","source_id":"s2"}],"assertion":"Copilot also gave his private phone number and full home address and offered a route there.","causal_attribution":"SWR as relayed by heise and ComputerBild (one chain); not in The Register."},{"id":"c3","status":"reported","evidence":[{"locator":"'Strafanzeige gestellt, ist damit aber abgeblitzt'; 'weil es keine reale Person gibt, die als Urheber zu betrachten ist.'","relation":"supports","source_id":"s1"},{"locator":"'had rejected criminal charges in two instances, and data protection officers could only achieve short-term success.'","relation":"supports","source_id":"s3"}],"assertion":"His criminal complaint was rejected (in two instances, he said) because no real person could be regarded as the author of the claims.","causal_attribution":"The reporter's account; the prosecutors' decisions were not inspected."},{"id":"c4","status":"reported","evidence":[{"locator":"'Als sich der zuständige Datenschutzbeauftragte vom Bayerischen Landesamt an Microsoft wendete, konnten die Beschuldigungen zunächst nicht mehr abgerufen werden. Einige Tage später antwortete der KI-Chatbot wohl aber wieder mit denselben falschen Behauptungen.'","relation":"supports","source_id":"s1"},{"locator":"'his lawyer has sent a cease-and-desist demand to Microsoft.'","relation":"supports","source_id":"s3"}],"assertion":"After the Bavarian data-protection authority contacted Microsoft the accusations were briefly unavailable, but a few days later the chatbot again gave the same false claims; his lawyer sent Microsoft a cease-and-desist demand.","causal_attribution":"SWR as relayed by heise (with its hedge wohl), and the reporter to The Register."},{"id":"c5","status":"reported","evidence":[{"locator":"'seeing his name associated with various crimes has been traumatizing'","relation":"supports","source_id":"s3"}],"assertion":"He said seeing his name associated with the crimes was traumatising.","causal_attribution":"The reporter's own account to The Register."},{"id":"c6","status":"reported","evidence":[{"locator":"'We investigated this report and have taken appropriate and immediate action to address it.'","relation":"supports","source_id":"s3"},{"locator":"'Aktuell sind die falschen Zuschreibungen nicht mehr abzurufen.'","relation":"supports","source_id":"s1"}],"assertion":"Microsoft said it had investigated the report and taken appropriate and immediate action; by late August 2024 the false answers were no longer returned.","causal_attribution":"Microsoft's statement to The Register; heise's own check at publication. Recorded as counterevidence and context."},{"id":"c7","status":"reported","evidence":[{"locator":"'Fast genau ein Jahr ist es jetzt her, dass der Journalist entdeckt hat, dass die Künstliche Intelligenz Copilot von Microsoft'","relation":"supports","source_id":"s4"}],"assertion":"He discovered the answers about a year before May 2025.","causal_attribution":"Pforzheimer Zeitung opening paragraph; basis for the 2024 event year."}],"effects":[{"label":"falsely described as a convicted child abuser and fraudster, with his contact details disclosed; said it was traumatising","claim_id":"c1","direction":"negative"}],"sources":[{"id":"s1","url":"https://www.heise.de/news/Copilot-macht-aus-einem-Gerichtsreporter-einen-Kinderschaender-9840437.html","kind":"news_report","access":"read","language":"de","translation_note":"Read in German by the research agent (an AI) on 2026-10-25 and translated by the research agent (an AI) into English; no human translator or reviewer was involved. heise online, 20 August 2024; HTTP 200, full text read. Relays the reporter's account to SWR; the SWR report itself was not fetched. The body names the reporter; this record does not.","independence_group":"swr-report-copilot-court-reporter-aug-2024"},{"id":"s2","url":"https://www.computerbild.de/artikel/cb-News-Internet-Microsoft-Copilot-macht-aus-einem-Journalisten-einen-verurteilten-Verbrecher-38844419.html","kind":"news_report","access":"read","language":"de","translation_note":"Read in German by the research agent (an AI) on 2026-10-25 and translated by the research agent (an AI) into English; no human translator or reviewer was involved. COMPUTER BILD, 20 August 2024; HTTP 200, full text read. Relays the SWR report; same origin as s1. Names the reporter.","independence_group":"swr-report-copilot-court-reporter-aug-2024"},{"id":"s3","url":"https://www.theregister.com/2024/08/26/microsoft_bing_copilot_ai_halluciation/","kind":"news_report","access":"read","language":"en","translation_note":"Read in English by the research agent (an AI) on 2026-10-25; no translation needed. The Register, 26 August 2024; HTTP 200, full text read. Adds the reporter's own email to The Register (the same witness as the SWR account) and a Microsoft spokesperson's statement. Names the reporter.","independence_group":"theregister-email-interview-aug-2024"},{"id":"s4","url":"https://www.pz-news.de/region_artikel,-KI-macht-PZ-Reporter-zu-einem-Verbrecher-Wer-traegt-dafuer-die-Verantwortung-_arid,2218448.html","kind":"news_report","access":"read","language":"de","translation_note":"Read in German by the research agent (an AI) on 2026-10-25 and translated by the research agent (an AI) into English; no human translator or reviewer was involved. Pforzheimer Zeitung, 17 May 2025; HTTP 200 but paywalled: only the opening paragraph was inspected, used for the year of discovery. Names the reporter.","independence_group":"pforzheimer-zeitung-may-2025"}],"version":1,"ai_roles":["own_use"],"contexts":["work","privacy","justice"],"unknowns":["On which dates and in which Copilot product version the answers were given, and how many other people saw them.","The prosecutors' decisions and the data-protection authority's file, which were not inspected.","Whether the cease-and-desist demand led to any legal proceedings or settlement."],"geography":{"basis":"He works as a court reporter in the Tübingen area for several publications (The Register). He found the answers by asking about himself; the country rests on his working area, not on the outlets or on the seat of the Bavarian data-protection authority that intervened.","court_countries":[],"event_countries":["DE"],"affected_person_countries":["DE"]},"publication":{"basis":"Published because Copilot's own answers attributed crimes to a named court reporter and, as reported, disclosed his contact details, and he described the effect; the account rests on his statements to SWR and The Register as reported, with Microsoft's statement.","reviewed_on":"2026-10-25"},"ai_involvement":{"basis":"Copilot's own answers made the claims: asked about the reporter, it attributed to him the crimes of defendants in trials he had covered, and, per SWR as relayed by heise and ComputerBild, gave his phone number, address and a route to it. The reporter found the answers by asking it about himself, and heise reports the answers returned after a temporary removal.","status":"supported"},"person_relations":["made_claim_about"]},"name":"German court reporter says Microsoft Copilot described him as a child abuser and fraudster and gave out his address; prosecutors rejected his complaint","summary":"A court reporter from the Tübingen area found in 2024 that Microsoft's Copilot, asked about him, described him as a man charged in a child abuse case, an escapee from a psychiatric institution and a fraudster who cheated widows, according to his account to the public broadcaster SWR as relayed by heise and ComputerBild. All were defendants in trials he had covered as a journalist. heise and ComputerBild, both citing SWR, report that Copilot also gave his private phone number and full home address and offered a route there. He told The Register that seeing his name tied to these crimes was traumatising and that prosecutors had rejected his criminal complaint in two instances; according to the SWR account relayed by heise, the complaint failed because no real person was the author. After the Bavarian data-protection authority contacted Microsoft the claims were briefly unavailable and then returned. Microsoft told The Register it had investigated and taken appropriate and immediate action.","incidentDate":"2024-01-01","incidentKind":"ongoing_experience","incidentDatePrecision":"year","exposurePattern":"repeated_interactions","reportedDate":"2024-08-20","aiSystem":"Microsoft Copilot (Bing Copilot), whose answers to questions about the reporter attributed crimes from his court reports to him","aiProduct":"Microsoft Copilot","severity":"medium","verificationStatus":"credible","harmCategories":[],"harmOutcomes":["reputational_harm","psychological_distress"],"harmOutcomeSummary":"Copilot attributed to a court reporter the crimes of defendants he had covered and, per SWR as relayed by heise and ComputerBild, gave out his phone number and address; he called the experience traumatising, and his criminal complaint was rejected (heise; The Register).","frameworkFacets":[],"causationStatus":"supported","participantUsersAffectedMin":1,"otherPeopleHarmedMin":0,"affectedCountStatus":"exact","affectedCountEvidence":"One person: the court reporter, who discovered the answers by asking Copilot about himself (The Register). Exact 1.","victimAgeRange":"adult","jurisdiction":"DE","platformType":"assistant","outcomeType":"internal_action","outcomeStatus":"unknown","primarySourceUrl":"https://www.heise.de/news/Copilot-macht-aus-einem-Gerichtsreporter-einen-Kinderschaender-9840437.html","primarySourceLabel":"heise online, 20 August 2024: Copilot turns a court reporter into a child abuser","firstPublishedAt":"2026-10-25T03:21:39.743612+00:00","updatedAt":"2026-10-25T03:21:39.743612+00:00","scopeVersion":"facts-v3","tags":["microsoft-copilot","false-accusation","defamation","personal-data-disclosure","germany","journalist"]},{"id":"2024-marana-arizona-gaggle-alert-on-pupils-unsent-draft-email-led-to-suspension-federal-court-lets-free-speech-claim-proceed","caseFacts":{"claims":[{"id":"c1","status":"reported","evidence":[{"locator":"Gaggle is a subscription, AI-based safety management software program that ... program that monitors students’ keystrokes on their laptops.","relation":"supports","source_id":"s2"},{"locator":"in October 2024, her son was drafting an email to a teacher at her home about a grade when he started typing out jokes. ... The threat was flagged by a monitoring software used by the school districts on their ,called Gaggle.","relation":"supports","source_id":"s3"},{"locator":"immediately deleted the message and closed his laptop, without","relation":"supports","source_id":"s1"}],"assertion":"According to the family's complaint, Gaggle, the AI-based safety management software the district ran on school laptops, monitors students' keystrokes, and it flagged a threat-like joke the pupil typed at home in October 2024 in a draft email to a teacher that he deleted without sending.","causal_attribution":"The family's allegation; the district declined to comment."},{"id":"c2","status":"reported","evidence":[{"locator":"Within an hour of ... School Resource ... in class, and they were concerned about a possible threat. ... reported having received a Gaggle notification regarding ... The Gaggle notification included a photograph of ... were being tracked or that anyone could see the draft email.","relation":"supports","source_id":"s1"}],"assertion":"The court's order recites the complaint's allegation that, within an hour, the principal and the school resource officer called the mother after the principal received a Gaggle notification that included a photograph of the pupil's screen showing the draft email, and that neither mother nor son knew his keystrokes were being tracked.","causal_attribution":"Allegations taken as true for the motion; not findings."},{"id":"c3","status":"reported","evidence":[{"locator":"would be suspended for ten days ‘for using a school device to write about ... suspended for forty-five days, to include the ten days currently being served, with the ... On December 5, 2024, ... suspension but modifying the length of ... suspension to nine days.","relation":"supports","source_id":"s1"}],"assertion":"According to the allegations the order recites, the pupil was suspended for ten days for using a school device to write about shooting up the school; a long-term suspension hearing set a 45-day suspension, reducible to eleven days with counselling; appeals affirmed it, and on 5 December 2024 the superintendent affirmed the suspension but shortened it to nine days.","causal_attribution":"The family's account of the district's disciplinary decisions."},{"id":"c4","status":"documented","evidence":[{"locator":"the Court will deny Defendants’ motion to ... dismiss regarding Plaintiffs’ First Amendment claim. ... the Court will dismiss his procedural due process ... to state a procedural due process claim arising from his long-term suspension. ... Dismiss (Doc. 20) is GRANTED in part and DENIED in part.","relation":"supports","source_id":"s1"}],"assertion":"In an order filed on 30 March 2026, the District of Arizona granted the district's motion to dismiss in part and denied it in part: it declined to dismiss the First Amendment claim and dismissed the procedural due process claims over the ten-day and long-term suspensions.","causal_attribution":"none"},{"id":"c5","status":"reported","evidence":[{"locator":"I had no idea that, down to the keystroke which was on the computer for a few seconds and deleted, that that would be seen ... MUSD declined to comment on this story, citing the pending litigation","relation":"supports","source_id":"s3"}],"assertion":"The pupil's mother told KOLD she had not known that a few seconds of deleted keystrokes could be seen; the district declined to comment, citing the pending litigation.","causal_attribution":"none"}],"effects":[{"label":"pupil suspended after an AI monitoring alert on an unsent, deleted draft email (family's allegations)","claim_id":"c3","direction":"negative"}],"sources":[{"id":"s1","url":"https://storage.courtlistener.com/recap/gov.uscourts.azd.1450446/gov.uscourts.azd.1450446.33.0.pdf","kind":"court_record","access":"read","language":"en","translation_note":"Read in English on 2026-10-23 (CourtListener RECAP copy of ECF 33, HTTP 200, all 17 pages); no translation. Order of the US District Court for the District of Arizona filed 30 March 2026 on the defendants' motion to dismiss; its factual background takes the first amended complaint's allegations as true for that purpose.","independence_group":"d-ariz-4-25-cv-00377-order"},{"id":"s2","url":"https://storage.courtlistener.com/recap/gov.uscourts.azd.1450446/gov.uscourts.azd.1450446.1.0.pdf","kind":"court_filing","access":"read","language":"en","translation_note":"Read in English on 2026-10-23 (CourtListener RECAP copy of ECF 1, the complaint filed 15 July 2025, HTTP 200); no translation. The first amended complaint (ECF 7) was not read; the order recites it.","independence_group":"family-complaint-d-ariz-2025"},{"id":"s3","url":"https://www.kold.com/2025/08/14/family-sues-marana-usd-after-student-was-suspended-school-shooting-joke/","kind":"news_report","access":"read","language":"en","translation_note":"Read in English on 2026-10-23 (HTTP 200); no translation. KOLD 13 News report of 14 August 2025 interviewing the pupil's mother and the family's lawyer; the district declined to comment. Same family account as the complaint.","independence_group":"family-complaint-d-ariz-2025"}],"version":1,"ai_roles":["institutional_use"],"contexts":["education","privacy","justice"],"unknowns":["The district's account of the alert and its decisions.","What category or score, if any, Gaggle assigned to the text, and whether a human reviewer at Gaggle saw it before the school.","How the case will be decided."],"geography":{"basis":"The pupil attended a school of the Marana Unified School District in Arizona, which issued the laptop, ran the monitoring and imposed the suspension; the suit is in the District of Arizona.","court_countries":["US"],"event_countries":["US"],"affected_person_countries":["US"]},"publication":{"basis":"Published as a contextual case (made_claim_about, informed_decision_about): according to the family's complaint, AI-based monitoring software flagged a pupil's deleted, unsent draft and the alert led to his suspension, and a federal court order of March 2026 lets the family's First Amendment claim proceed. The facts are the family's allegations; the district has not commented. The pupil is not identified.","reviewed_on":"2026-10-23"},"ai_involvement":{"basis":"The family's complaint describes Gaggle as AI-based safety management software that monitors students' keystrokes, and alleges, as the court's order recites, that Gaggle sent the principal a notification with a photograph of the pupil's screen showing a deleted, unsent draft containing a threat-like joke. That automated flag of his text as concerning content is what brought the draft to the school's attention: the principal and school resource officer called within the hour, and the district suspended him for writing it. The district has not commented, and how the software classified the text is not described beyond the notification.","status":"reported"},"person_relations":["made_claim_about","informed_decision_about"]},"name":"Arizona: family says a Gaggle alert on a pupil's unsent joke email led to his suspension; federal court lets their free-speech claim proceed","summary":"On 9 October 2024 a Marana Unified School District pupil in Arizona, typing at home on a school Chromebook in front of his mother, wrote a threat-like joke about shooting up the school in a draft email to a teacher and deleted it without sending it, according to his family's federal lawsuit. The family says Gaggle, the AI-based monitoring software the district used, flagged it: the principal received a Gaggle notification with a photograph of the screen and, with the school resource officer, called the mother within the hour. The family says the pupil was suspended for ten days, a long-term suspension hearing imposed 45 days reducible to eleven with counselling, and the superintendent finally affirmed a nine-day suspension in December 2024. In March 2026 the District of Arizona let the family's First Amendment claim proceed and dismissed its procedural due process claims. The district declined to comment to the press.","incidentDate":"2024-10-09","incidentEndDate":"2024-12-05","incidentKind":"single_event","incidentDatePrecision":"day","exposurePattern":"single_interaction","reportedDate":"2025-07-15","aiSystem":"Gaggle, described in the family's complaint as subscription, AI-based safety management software that monitors students' keystrokes on school laptops; its notification to the principal included a photograph of the pupil's screen","aiProduct":"Gaggle (reported)","severity":"low","verificationStatus":"credible","harmCategories":[],"harmOutcomes":["educational_harm"],"harmOutcomeSummary":"According to his family's lawsuit, a Gaggle alert on a deleted, unsent joke in a draft email led a school district to suspend a pupil; the superintendent finally set nine days. A federal court let the family's free-speech claim proceed.","frameworkFacets":[],"pathwayDirection":"outward","causationStatus":"alleged","participantUsersAffectedMin":0,"otherPeopleHarmedMin":1,"affectedCountStatus":"exact","affectedCountEvidence":"One person: the pupil suspended after the Gaggle alert (complaint, order and KOLD). His parents are plaintiffs on his behalf and are not counted separately.","victimAgeRange":"minor","jurisdiction":"US","platformType":"other","outcomeType":"lawsuit_ongoing","outcomeStatus":"ongoing","primarySourceUrl":"https://storage.courtlistener.com/recap/gov.uscourts.azd.1450446/gov.uscourts.azd.1450446.33.0.pdf","primarySourceLabel":"US District Court, District of Arizona, No. 4:25-cv-00377, order on motion to dismiss filed 30 Mar 2026 (ECF 33)","firstPublishedAt":"2026-10-23T03:21:47.656673+00:00","updatedAt":"2026-10-23T03:21:47.656673+00:00","scopeVersion":"facts-v3","tags":["school","student-monitoring","gaggle","suspension","first-amendment","lawsuit"]},{"id":"2024-ontario-hospital-otter-ai-bot-recorded-physicians-rounds-meeting-and-emailed-patient-health-information","caseFacts":{"claims":[{"id":"c1","status":"documented","evidence":[{"locator":"Letter: 'October 27, 2025'; 'Reported Breach HR24-00691'; 'On December 17, 2024, you reported a privacy breach under the Personal Health Information Protection Act'.","relation":"supports","source_id":"s1"}],"assertion":"The Information and Privacy Commissioner of Ontario wrote to the hospital on 27 October 2025 about a privacy breach the hospital had reported on 17 December 2024 under the Personal Health Information Protection Act (file HR24-00691).","causal_attribution":"The letter itself establishes its date, file and the reported breach it addresses."},{"id":"c2","status":"reported","evidence":[{"locator":"Letter: 'a virtual hepatology rounds meeting attended by hospital physicians was inadvertently recorded by an artificial intelligence-powered transcription tool'; 'The physician left his employment with the hospital in June 2023. He installed Otter.ai on his personal device in September 2024.'; 'the Tool was able to join the September 23, 2024 hepatology rounds meeting via accessing the link to it in the physician'; 'and record the meeting without' (page break) 'notice. This was discovered'.","relation":"supports","source_id":"s1"},{"locator":"'an Otter.ai bot attended the meeting in his place'","relation":"supports","source_id":"s2"}],"assertion":"According to the hospital's report as recorded by the IPC, on 23 September 2024 the Otter.ai AI transcription tool joined a virtual hepatology rounds meeting of hospital physicians through the meeting link in a former physician's personal calendar and recorded the meeting without notice; the physician had left the hospital in June 2023 and installed Otter.ai on his personal device in September 2024.","causal_attribution":"The hospital's own breach report, relayed by the regulator; the Globe relays the same letter."},{"id":"c3","status":"reported","evidence":[{"locator":"Letter: 'the personal health information of seven patients was discussed amongst the meeting participants and captured in the transcript'; 'diagnoses, medical notes, and treatment information'.","relation":"supports","source_id":"s1"}],"assertion":"The personal health information of seven patients admitted to the hospital, including names, sex, diagnoses, medical notes and treatment information, was discussed in the meeting and captured in the transcript.","causal_attribution":"Hospital's report as recorded by the IPC."},{"id":"c4","status":"reported","evidence":[{"locator":"Letter: 'This was discovered when a meeting summary and access to a transcript of the recording was automatically emailed to participants after the meeting'; 'of the 65 users on the recipient list, 53 responded confirming they either deleted the email or never received it'; 'the remaining 12 individuals appear to have departed from the hospital'.","relation":"supports","source_id":"s1"},{"locator":"'sent a summary and a link to a transcript of the meeting to 65 people on the invite list, 12 of whom were no longer employed by the hospital'","relation":"supports","source_id":"s2"}],"assertion":"The breach was discovered when the tool automatically emailed a meeting summary and access to the transcript to participants; 65 users were on the recipient list, 53 confirmed deleting the email or never receiving it, and the remaining 12 appeared to have left the hospital.","causal_attribution":"Hospital's report as recorded by the IPC; the Globe relays the same figures."},{"id":"c5","status":"reported","evidence":[{"locator":"Letter: 'the physician did not respond to this request'; 'Submit a formal request to Otter.ai to delete any personal health information'.","relation":"supports","source_id":"s1"}],"assertion":"The hospital asked the former physician to have Otter.ai delete the recording, but he did not respond; the IPC recommended that the hospital itself formally ask Otter.ai to delete any patient information retained from the meeting.","causal_attribution":"Mixed origin: the recommendation is the IPC's own and documented by the letter; the physician's non-response is the hospital's report recorded in the letter, so the whole claim is reported."},{"id":"c6","status":"reported","evidence":[{"locator":"Letter: 'five of seven affected individuals were notified of the breach by letter on December 17, 2024'; 'the other two affected individuals are deceased'; 'For one of these individuals, the hospital provided a copy of the notice letter to their estate trustee'; 'uploaded a copy of the notice letter to the patient’s electronic health record'.","relation":"supports","source_id":"s1"}],"assertion":"Five of the seven patients were notified by letter on 17 December 2024; the other two had died, and the hospital sent a notice to one estate trustee and placed a copy in the other patient's health record.","causal_attribution":"Hospital's report as recorded by the IPC; the deaths are not connected to the breach in any source."}],"effects":[{"label":"patients' health information was recorded by an unapproved AI transcription bot and emailed to a recipient list that included former staff","claim_id":"c3","direction":"negative"}],"sources":[{"id":"s1","url":"https://www.ipc.on.ca/en/media/5995/download","kind":"official_record","access":"read","language":"en","translation_note":"Letter of the Information and Privacy Commissioner of Ontario to the hospital, Reported Breach HR24-00691, dated 27 October 2025, released without the hospital's name; the ipc.on.ca download returned HTTP 403 to this fetcher and the full letter was read from an Internet Archive capture on 2026-10-21.","independence_group":"ipc-letter-hr24-00691"},{"id":"s2","url":"https://www.theglobeandmail.com/canada/article-ai-bot-doctors-meeting-patient-info-hospital-privacy-watchdog/","kind":"news","access":"read","language":"en","translation_note":"The Globe and Mail, 19 November 2025; read in full on 2026-10-21. Its account of the breach comes from the IPC letter, so it shares that letter's group.","independence_group":"ipc-letter-hr24-00691"},{"id":"s3","url":"https://www.hrreporter.com/focus-areas/automation-ai/ai-bot-sends-confidential-info-to-ontario-hospital-patients-after-recording-doctors-meeting-ipc/393766","kind":"news","access":"read","language":"en","translation_note":"Canadian HR Reporter, 21 November 2025; read in full on 2026-10-21. Its account comes from the IPC letter; its headline says information went to patients, which its own text and the letter do not support.","independence_group":"ipc-letter-hr24-00691"}],"version":1,"ai_roles":["others_use"],"contexts":["privacy","health","work"],"unknowns":["Whether Otter.ai deleted the recording and transcript; the hospital had no confirmation when the letter was written.","Whether the 12 departed recipients deleted the email.","Whether the former physician knew the tool had joined the meeting.","What the hospital reported to the IPC by the 27 January 2026 update deadline."],"geography":{"basis":"The letter is from Ontario's privacy commissioner to a hospital custodian under Ontario's health privacy law, and the patients had been admitted to that hospital; the Globe describes it as an Ontario hospital. No court is involved.","court_countries":[],"event_countries":["CA"],"affected_person_countries":["CA"]},"publication":{"basis":"Published under the 2026-10-06 privacy clarification: a regulator's letter records the hospital's report that an AI transcription tool joined a clinical meeting without notice, captured seven patients' health information and emailed it to a list including former staff. The relation is received_personal_information.","reviewed_on":"2026-10-21"},"ai_involvement":{"basis":"The hospital's report recorded in the regulator's letter says the Otter.ai tool itself joined the rounds meeting from a former physician's calendar, recorded it without notice, captured seven patients' health information in its transcript and automatically emailed a summary and transcript access to 65 recipients. That capture and distribution by the AI service is the privacy harm.","status":"supported"},"person_relations":["received_personal_information"]},"name":"Ontario hospital reports that an Otter.ai bot joined physicians' rounds through a former doctor's invite, recorded them and emailed patient information","summary":"An Ontario hospital reported to the provincial privacy commissioner that on 23 September 2024 the Otter.ai transcription tool, installed on a personal device by a physician who had left the hospital in June 2023 and was still on the meeting invite, joined a virtual hepatology rounds meeting, recorded it without notice and then emailed a summary and transcript access to the invite list. The personal health information of seven patients, including diagnoses and treatment, was captured; 12 of the 65 recipients appeared to have left the hospital. The commissioner's letter of 27 October 2025 recommended that the hospital itself ask Otter.ai to delete the retained patient data.","incidentDate":"2024-09-23","incidentKind":"single_event","incidentDatePrecision":"day","exposurePattern":"single_interaction","reportedDate":"2025-11-19","aiSystem":"Otter.ai, an AI transcription tool that a former hospital physician installed on a personal device and that joined meetings from his calendar (per the hospital's report to the IPC)","aiProduct":"Otter.ai (reported)","aiCompany":"Otter.ai","severity":"medium","verificationStatus":"credible","harmCategories":[],"harmOutcomes":["other_material_harm"],"harmOutcomeSummary":"Seven patients' health information was recorded by an unapproved AI transcription bot, which then emailed a summary and transcript access to 65 people, 12 of whom appeared to have left the hospital, according to the hospital's report to Ontario's privacy commissioner.","frameworkFacets":[],"causationStatus":"supported","participantUsersAffectedMin":0,"otherPeopleHarmedMin":7,"affectedCountStatus":"exact","affectedCountEvidence":"The IPC letter records that the personal health information of seven patients was discussed in the meeting and captured in the transcript. Meeting participants and email recipients are not counted as harmed.","victimAgeRange":"unknown","jurisdiction":"CA-ON","platformType":"assistant","outcomeType":"regulatory_action","outcomeStatus":"resolved","primarySourceUrl":"https://www.ipc.on.ca/en/media/5995/download","primarySourceLabel":"Information and Privacy Commissioner of Ontario, letter on Reported Breach HR24-00691, 27 October 2025","firstPublishedAt":"2026-10-21T03:12:57.400852+00:00","updatedAt":"2026-10-21T03:12:57.400852+00:00","scopeVersion":"facts-v3","tags":["otter-ai","ai-notetaker","meeting-transcription","health-information","privacy-breach","regulator","ontario"]},{"id":"2024-us-otter-notetaker-recorded-non-users-meetings-privacy-suit","caseFacts":{"claims":[{"id":"c1","status":"reported","evidence":[{"locator":"'participated in a Zoom meeting in California on February'; '24, 2025, where the Otter Notetaker was used by a meeting participant to transcribe the'; 'participated in a Zoom meeting in California in March 2025'; 'used Zoom to communicate with a medical professional, and'; 'Illinois on January 10, 2025 and May 19, 2025 where the Otter Notetaker was used to'; 'participated in a Zoom meeting in Washington state in March'","relation":"supports","source_id":"s1"},{"locator":"'Plaintiffs are seven individuals from California, Illinois, and Washington state who'; 'participated in virtual meetings where the Otter Notetaker was allegedly used without their'; 'consent to record, transcribe, and store the contents of their communications.'","relation":"supports","source_id":"s2"},{"locator":"'Otter lacks knowledge or information sufficient to form a belief as to the truth of the'; 'allegations concerning the named Plaintiff’s personal knowledge, experiences, expectations, or'","relation":"contradicts","source_id":"s3"},{"locator":"'who alleges his privacy was \"severely invaded\" upon realizing Otter was secretly recording a confidential conversation.'","relation":"supports","source_id":"s5"}],"assertion":"Seven people who were not Otter account holders allege that, on Zoom or Microsoft Teams calls (six give dates between March 2024 and May 2025), another participant used Otter Notetaker and Otter recorded, transcribed and stored their conversations without their consent; the calls include one with a medical professional and meetings with a financial professional.","causal_attribution":"The plaintiffs attribute the recording to Otter's product design, which they say lets the notetaker record without every participant's consent; Otter denies the allegations."},{"id":"c2","status":"reported","evidence":[{"locator":"'other personal information for its own commercial use, including to train its automatic speech'; 'The Illinois Plaintiffs allege that, during their virtual meetings, Otter'","relation":"supports","source_id":"s2"},{"locator":"'Otter claims that before the audio of meetings is fed into its machine learning systems to help improve an AI speech recognition feature, it is \"de-identified,\"'","relation":"context","source_id":"s5"}],"assertion":"The plaintiffs allege that Otter keeps their conversational data on its servers and uses it to train its speech-recognition and machine-learning models, and that it captured the Illinois plaintiffs' voiceprints; Otter's policy, as reported by NPR, says it trains on transcripts with users' permission and de-identifies audio.","causal_attribution":"Alleged by the plaintiffs; Otter denies the allegations."},{"id":"c3","status":"reported","evidence":[{"locator":"'felt frustrated, embarrassed, and stressed to learn that his'; 'conversation was recorded without his consent, and his information, voice, and'","relation":"supports","source_id":"s1"}],"assertion":"One California plaintiff alleges that he felt frustrated, embarrassed and stressed to learn that his conversation had been recorded without his consent.","causal_attribution":"Alleged by the plaintiffs as a consequence of Otter's recording."},{"id":"c4","status":"documented","evidence":[{"locator":"'Accordingly, the motion to dismiss Plaintiffs’ claims for lack of standing is denied.'; 'Accordingly, the motion to dismiss Plaintiffs’ ECPA claim is denied.'; 'Accordingly, the motion to dismiss Plaintiffs’ BIPA claims for lack of standing is denied.'; 'The motion to dismiss is GRANTED with leave to amend as to Count 2'; 'The motion is DENIED in all other respects.'","relation":"supports","source_id":"s2"}],"assertion":"On 13 August 2026 the court denied Otter's motion to dismiss for lack of standing, holding that the alleged interception and retention of private conversations was a concrete injury, let the federal wiretap, California eavesdropping and Illinois biometric claims proceed, and dismissed some other claims.","causal_attribution":"Procedural ruling on the pleadings; no finding that Otter recorded anyone unlawfully."}],"effects":[{"label":"the plaintiffs say Otter's notetaker recorded, transcribed and stored their conversations on calls they joined, without their consent, including a medical call and calls with a financial professional","claim_id":"c1","direction":"negative"},{"label":"the plaintiffs say their conversations stay on Otter's servers for training its models and that it captured voiceprints of the Illinois plaintiffs","claim_id":"c2","direction":"negative"},{"label":"one plaintiff says he felt frustrated, embarrassed and stressed to learn he had been recorded","claim_id":"c3","direction":"negative"}],"sources":[{"id":"s1","url":"https://storage.courtlistener.com/recap/gov.uscourts.cand.454675/gov.uscourts.cand.454675.35.0.pdf","kind":"court_filing","access":"read","language":"en","translation_note":"Consolidated Class Action Complaint, ECF 35, filed 5 December 2025; RECAP PDF (68 pages) read on 2026-09-29 (HTTP 200), plaintiffs' experience sections read in full. Plaintiffs' allegations, not findings.","independence_group":"plaintiffs-account"},{"id":"s2","url":"https://storage.courtlistener.com/recap/gov.uscourts.cand.454675/gov.uscourts.cand.454675.68.0.pdf","kind":"court_filing","access":"read","language":"en","translation_note":"Order Granting Motion to Dismiss in Part, ECF 68, 13 August 2026 (Judge Eumi K. Lee); RECAP PDF read in full on 2026-09-29. Rules on the pleadings and accepts the allegations as true for that purpose; it makes no finding of fact.","independence_group":"ndcal-court-record"},{"id":"s3","url":"https://storage.courtlistener.com/recap/gov.uscourts.cand.454675/gov.uscourts.cand.454675.80.0.pdf","kind":"court_filing","access":"read","language":"en","translation_note":"Otter.ai's Answer and Affirmative Defenses to Plaintiffs' Consolidated Class Action Complaint, ECF 80, 17 September 2026; RECAP PDF read on 2026-09-29, plaintiff-experience paragraphs and responses read.","independence_group":"otter-answer"},{"id":"s4","url":"https://storage.courtlistener.com/recap/gov.uscourts.cand.454675/gov.uscourts.cand.454675.1.0.pdf","kind":"court_filing","access":"read","language":"en","translation_note":"Brewer v. Otter.ai, Inc., original Class Action Complaint, ECF 1, filed 15 August 2025; RECAP PDF read on 2026-09-29.","independence_group":"plaintiffs-account"},{"id":"s5","url":"https://www.npr.org/2025/08/15/g-s1-83087/otter-ai-transcription-class-action-lawsuit","kind":"news_report","access":"read","language":"en","translation_note":"NPR, 15 August 2025, read live on 2026-09-29 (HTTP 200). Reports the original complaint and summarises Otter's privacy policy; says neither side responded to requests for comment.","independence_group":"plaintiffs-account"},{"id":"s6","url":"https://storage.courtlistener.com/recap/gov.uscourts.cand.454675/gov.uscourts.cand.454675.79.0.pdf","kind":"court_filing","access":"read","language":"en","translation_note":"Amended Case Management and Scheduling Order, ECF 79, 16 September 2026; read on 2026-09-29. Procedural facts only.","independence_group":"ndcal-court-record"}],"version":1,"ai_roles":["others_use"],"contexts":["privacy","work","health","finance"],"unknowns":["Whether Otter's product notified the plaintiffs on their calls that the notetaker was present, and what the other participants' Otter settings were.","Whether the plaintiffs' conversations were in fact used to train Otter's models, and in what form.","The exact date of one Illinois plaintiff's meetings, which the complaint places only within the limitations period."],"geography":{"basis":"The consolidated complaint places the plaintiffs' calls in California, in Chicago and elsewhere in Illinois, and in Washington state; the order describes the plaintiffs as individuals from California, Illinois and Washington state. The case is in the U.S. District Court for the Northern District of California.","court_countries":["US"],"event_countries":["US"],"affected_person_countries":["US"]},"publication":{"basis":"Published as a bounded series of concrete privacy harms reported by identified people in a pending federal suit: each named plaintiff describes a call on which another participant's AI notetaker recorded and transcribed them without consent. The account rests on the plaintiffs' pleadings and the court's ruling on them; Otter denies the allegations. Plaintiff names are omitted.","reviewed_on":"2026-10-13"},"ai_involvement":{"basis":"Each plaintiff alleges that another participant's Otter account joined the Otter Notetaker, an AI transcription assistant, to a Zoom or Microsoft Teams call the plaintiff was on, and that the notetaker recorded, transcribed and stored the plaintiff's conversation on Otter's servers without the plaintiff's knowledge or consent (ECF 35, the consolidated complaint; the court's 13 August 2026 order summarises the same allegation). That alleged capture of the plaintiffs' conversations by an AI service is the reported harm. The complaint also alleges that Otter used the recordings to train its speech-recognition models and took voiceprints of the Illinois plaintiffs. Otter admits it offers an AI meeting assistant called Otter Notetaker but says it lacks knowledge of the plaintiffs' calls and denies the allegations. The capture is alleged, not established, so the status is reported.","status":"reported"},"person_relations":["received_personal_information"]},"name":"United States: seven people who say they had no Otter account allege its AI Notetaker recorded calls without consent; a judge let core privacy claims proceed","summary":"Seven people in California, Illinois and Washington state are suing Otter.ai in a consolidated class action in the Northern District of California. Each says another meeting participant used Otter's AI Notetaker on a Zoom or Microsoft Teams call they joined, and that Otter recorded, transcribed and stored their conversation without their consent; six of them date their calls between March 2024 and May 2025. They say they were not Otter account holders. The calls they describe include a medical consultation, meetings with a financial professional and work discussions. They allege that Otter keeps the recordings and uses them to train its speech-recognition models, that it captured voiceprints of the Illinois plaintiffs, and one of them says learning of the recording left him frustrated, embarrassed and stressed. On 13 August 2026 the court found that the alleged interception of private conversations was a concrete injury and let the federal wiretap, California eavesdropping and Illinois biometric claims proceed, while dismissing some claims. Otter's answer to the consolidated complaint, filed on 17 September 2026, says it lacks knowledge of the plaintiffs' experiences and denies the allegations.","incidentDate":"2024-03-01","incidentEndDate":"2025-05-19","incidentKind":"bounded_series","incidentDatePrecision":"month","exposurePattern":"single_interaction","reportedDate":"2025-08-15","aiSystem":"Otter Notetaker, Otter.ai's AI meeting assistant, joined to Zoom and Microsoft Teams calls by another participant's Otter account","aiProduct":"Otter Notetaker","aiCompany":"Otter.ai, Inc.","severity":"low","verificationStatus":"credible","harmCategories":[],"harmOutcomes":["other_material_harm","psychological_distress"],"harmOutcomeSummary":"The plaintiffs allege that Otter recorded, transcribed and kept their private conversations on calls they joined without their consent, including medical and financial discussions, used them to train its models and took voiceprints, and one plaintiff says learning this left him frustrated, embarrassed and stressed (plaintiffs' allegations in a pending suit; Otter denies them).","frameworkFacets":[],"causationStatus":"alleged","participantUsersAffectedMin":0,"otherPeopleHarmedMin":7,"affectedCountStatus":"documented_minimum","affectedCountEvidence":"The court order says 'Plaintiffs are seven individuals from California, Illinois, and Washington state' who participated in meetings where the Otter Notetaker was allegedly used without their consent; each describes their own call in the consolidated complaint. They were not Otter account holders (one says only that to her knowledge she did not create an account), so they are counted as other people. The proposed class members are not counted. Documented minimum 7.","victimAgeRange":"adult","jurisdiction":"US","platformType":"assistant","outcomeType":"lawsuit_ongoing","outcomeStatus":"ongoing","primarySourceUrl":"https://storage.courtlistener.com/recap/gov.uscourts.cand.454675/gov.uscourts.cand.454675.68.0.pdf","primarySourceLabel":"Order granting motion to dismiss in part, In re Otter.AI Privacy Litigation, No. 5:25-cv-06911-EKL (N.D. Cal.), 13 August 2026","firstPublishedAt":"2026-09-29T09:04:08.493618+00:00","updatedAt":"2026-10-13T03:15:43.76525+00:00","scopeVersion":"facts-v3","tags":["otter-ai","ai-notetaker","meeting-transcription","privacy","wiretap","bipa","voiceprint","training-data","class-action","non-user"]},{"id":"2025-holmen-norway-gdpr","caseFacts":{"claims":[{"id":"c1","status":"documented","evidence":[{"locator":"'Vienna, 20.03.2025'; 'noyb is representing the complainant under Article 80(1) GDPR'; '58(2)(d) GDPR to order the respondent to delete the defamatory output on the'","relation":"supports","source_id":"s2"},{"locator":"'noyb has therefore filed a complaint with the Norwegian Datatilsynet'; 'noyb is asking the Datatilsynet to order OpenAI to delete the defamatory output and fine-tune its model to eliminate inaccurate results'","relation":"supports","source_id":"s1"}],"assertion":"On 20 March 2025 noyb filed a complaint against OpenAI with Norway's data protection authority, Datatilsynet, representing Holmen, and asked it to order OpenAI to delete the defamatory output and fine-tune its model, and suggested a fine.","causal_attribution":"The filing itself establishes that the complaint was made and what it asks; it does not establish its allegations."},{"id":"c2","status":"reported","evidence":[{"locator":"'Arve Hjalmar Holmen was accused and later convicted of murdering his'; 'two sons, as well as for the attempted murder of his third son.'","relation":"supports","source_id":"s2"},{"locator":"'A Norwegian man has filed a complaint after ChatGPT falsely told him he had killed two of his sons and been jailed for 21 years.'","relation":"supports","source_id":"s3"},{"locator":"'NRK har sett en skjermdump av chatten mellom mannen og ChatGPT.'; 'ChatGPT hevdet videre at mannen ble dømt til 21 års fengsel'","relation":"supports","source_id":"s4"}],"assertion":"According to the complaint, when he asked ChatGPT \"Who is Arve Hjalmar Holmen?\", it replied that he had been accused and convicted of murdering two of his sons and of attempting to murder his third son, and sentenced to 21 years in prison.","causal_attribution":"The reply text comes from the complainant's screenshot as quoted in the complaint; NRK says it saw the screenshot. The full conversation was not published, and OpenAI did not dispute that the answer was given."},{"id":"c3","status":"reported","evidence":[{"locator":"'he has never been accused nor convicted of any crime and is'","relation":"supports","source_id":"s2"},{"locator":"'the fake story included real elements of his personal life. Among them were the actual number and the gender of his children and the name of his home town.'","relation":"supports","source_id":"s1"}],"assertion":"The complaint says the story was false and that he has never been accused or convicted of any crime, while the reply contained real details about him: his home town and the number and gender of his children.","causal_attribution":"The complainant's account through noyb. No source disputes that the crimes are fictitious."},{"id":"c4","status":"reported","evidence":[{"locator":"'The complainant was deeply troubled by these outputs, which could have harmful'","relation":"supports","source_id":"s2"},{"locator":"'The fact that someone could read this output and believe it is true, is what scares me the most.'","relation":"supports","source_id":"s1"},{"locator":"'the fact that someone could read this output and believe it is true is what scares me the most'","relation":"supports","source_id":"s3"},{"locator":"'The complainant was deeply troubled by these outputs, which could have harmful effect in his private life'","relation":"supports","source_id":"s5"}],"assertion":"The complaint says he was deeply troubled by the output, which could harm his private life if reproduced in his community; he said that the thought that someone could read it and believe it scared him most.","causal_attribution":"His own statements, issued through noyb and repeated by the BBC; one origin. The complaint does not report that anyone else saw the output."},{"id":"c5","status":"reported","evidence":[{"locator":"'output, however OpenAI responded with a “template-answer” and not with a tailored'","relation":"supports","source_id":"s2"}],"assertion":"The complaint says he contacted OpenAI about the false output and received a template answer rather than a reply to his request.","causal_attribution":"The complaint's allegation; the annexes were not published."},{"id":"c6","status":"reported","evidence":[{"locator":"'ChatGPT has changed its model since Mr Holmen's search in August 2024'; 'Noyb told the BBC Mr Holmen had made a number of searches that day'; 'They also acknowledged the previous searches could have influenced the answer about his children'","relation":"supports","source_id":"s3"}],"assertion":"The BBC reports that he made the search in August 2024; noyb told the BBC that he made several searches that day, including his brother's name, which produced multiple incorrect stories, and acknowledged that the earlier searches could have influenced the answer about his children.","causal_attribution":"Single source for the date and the context; the complaint itself redacts the date."},{"id":"c7","status":"reported","evidence":[{"locator":"'According to noyb, ChatGPT's memory function, which can recall personal information from previous chats, was turned off during this interaction.'; 'Our own verification attempts using ChatGPT, Mistral, Google Gemini, Deepseek, and Claude haven't been able to reproduce the false claims.'","relation":"supports","source_id":"s6"}],"assertion":"The Decoder reports that noyb said ChatGPT's memory function was off during the exchange and that he had other conversations in the same chat, that only a screenshot was released, and that its own attempts with several chatbots did not reproduce the claims.","causal_attribution":"Counterevidence on how the answer arose; it leaves open whether earlier prompts in the chat shaped it."},{"id":"c8","status":"reported","evidence":[{"locator":"'While we're still reviewing this complaint, it relates to a version of ChatGPT which has since been enhanced with online search capabilities that improves accuracy.'","relation":"supports","source_id":"s3"},{"locator":"'ChatGPT has stopped telling lies about him being a murderer. However, the incorrect data may still remain part of the LLM’s dataset.'","relation":"supports","source_id":"s1"}],"assertion":"OpenAI said it was still reviewing the complaint and that it related to a version of ChatGPT since enhanced with online search; noyb said ChatGPT had since stopped describing him as a murderer but that the incorrect data may remain in the model.","causal_attribution":"Each party's own statement, attributed."},{"id":"c9","status":"reported","evidence":[{"locator":"'Det norske datatilsynet sier at de mottok klagen i dag og ikke har rukket å saksbehandle den.'; 'Det mest sannsynlige er at denne klagen må sendes videre til det europeiske datatilsynet'","relation":"supports","source_id":"s4"}],"assertion":"Datatilsynet confirmed to NRK that it received the complaint on 20 March 2025, and its international section head said it would most likely have to be forwarded to another European data protection authority.","causal_attribution":"The regulator's statement as reported by NRK; no later decision was found."}],"effects":[{"label":"distress at a fabricated account that he murdered his children, as the complaint and his statement describe","claim_id":"c4","direction":"negative"}],"sources":[{"id":"s1","url":"https://noyb.eu/en/ai-hallucinations-chatgpt-created-fake-child-murderer","kind":"press_release","access":"read","language":"en","translation_note":"Read live on 2026-11-01 (HTTP 200; bodies/lr-holmen-noyb.html/.txt). noyb press release announcing the complaint; same origin as the complaint.","independence_group":"noyb-complaint-20250320"},{"id":"s2","url":"https://noyb.eu/sites/default/files/2025-03/OpenAI_complaint_redacted.pdf","kind":"complaint_filing","access":"read","language":"en","translation_note":"Read on 2026-11-01 (HTTP 200; PDF text extracted with pdftotext to bodies/lr-holmen-complaint.txt). Redacted complaint; the query date, residence and the conversation link are blacked out. Annexes were not published.","independence_group":"noyb-complaint-20250320"},{"id":"s3","url":"https://www.bbc.com/news/articles/c0kgydkr516o","kind":"news_report","access":"read","language":"en","translation_note":"Read live on 2026-11-01 (HTTP 200; bodies/lr-holmen-bbc.html/.txt). Quotes the reply from the complaint, and adds an OpenAI statement and noyb remarks to the BBC (date of the search, other searches that day).","independence_group":"bbc-20250321"},{"id":"s4","url":"https://www.nrk.no/norge/nordmann-klager-chatgpt-til-datatilsynet-1.17348958","kind":"news_report","access":"read","language":"no","translation_note":"Read live on 2026-11-01 (HTTP 200; bodies/lr-holmen-nrk.html/.txt). NRK says it saw a screenshot of the chat and quotes Datatilsynet; it did not reach him and quotes him from Aftenposten (paywalled, not read). Norwegian; quoted passages translated by the research agent (an AI) for this review; no human translator checked them.","independence_group":"nrk-20250320"},{"id":"s5","url":"https://fortune.com/2025/03/21/chatgpt-murder-hallucination-arve-hjalmar-holmen-noyb-openai-complaint","kind":"news_report","access":"read","language":"en","translation_note":"Read live on 2026-11-01 (HTTP 200; bodies/lr-holmen-fortune.html/.txt). Relays the complaint (grouped with it) and adds lawyers' comments; no independent account of the event.","independence_group":"noyb-complaint-20250320"},{"id":"s6","url":"https://the-decoder.com/chatgpts-bizarre-child-murder-claims-about-arve-hjalmar-holmen-leave-some-questions-unresolved/","kind":"news_report","access":"read","language":"en","translation_note":"Read live on 2026-11-01 (HTTP 200; bodies/lr-holmen-decoder.html/.txt). Reports noyb statements about the chat context and its own failed attempts to reproduce the answer.","independence_group":"decoder-20250321"}],"version":1,"ai_roles":["own_use"],"contexts":["privacy","everyday_life"],"unknowns":["The full conversation, including the earlier prompts that day that noyb acknowledged could have influenced the answer.","The exact date of the search: the complaint redacts it and the BBC gives August 2024.","Where he made the search.","The outcome of the complaint: no decision by Datatilsynet or any authority it may have forwarded the complaint to was found by 1 November 2026.","Whether anyone other than him saw the output."],"geography":{"basis":"Carried forward from the 2026-09-23 metadata review and re-checked: noyb and the complaint describe him as a Norwegian user and citizen, so the affected-person country is NO. No body says where he used ChatGPT, so the event country stays a reviewed unknown. The complaint went to a regulator, Datatilsynet, and no court is involved, so court_countries is empty.","court_countries":[],"event_countries":[],"affected_person_countries":["NO"]},"publication":{"basis":"Published as a case of a chatbot making a false claim about a person to that person (made_claim_about and communicated_with). The complaint quotes ChatGPT's reply stating that he murdered two of his sons, and records his distress. The claims are attributed to the complaint, his statements and the parties, and the counterevidence on earlier prompts is kept.","reviewed_on":"2026-11-01"},"ai_involvement":{"basis":"The complaint quotes ChatGPT's reply to his question \"Who is Arve Hjalmar Holmen?\": it said he had been convicted of murdering two of his sons and attempting to murder a third, and sentenced to 21 years. NRK says it saw a screenshot of the chat, and OpenAI attributed the answer to an earlier ChatGPT version without disputing it. The complaint connects his distress to that generated claim about him. Earlier prompts in the same chat may have shaped the answer, as noyb acknowledged.","status":"reported"},"person_relations":["made_claim_about","communicated_with"]},"name":"Norwegian man's GDPR complaint says ChatGPT answered a question about him with a fabricated story that he had murdered two of his sons","summary":"Arve Hjalmar Holmen asked ChatGPT \"Who is Arve Hjalmar Holmen?\" in August 2024, according to the BBC. The digital-rights group noyb's complaint on his behalf quotes the reply: it said he had been convicted of murdering two of his sons and of attempting to murder a third, and sentenced to 21 years in prison. The complaint says none of this happened and that he has never been accused or convicted of any crime, while the reply did contain real details such as his home town and the number and gender of his children. The complaint says he was deeply troubled by the output, and he said the thought that someone could read it and believe it scared him most. noyb filed the complaint with Norway's data protection authority, Datatilsynet, on 20 March 2025, asking it to order OpenAI to delete the output and correct its model and to impose a fine. OpenAI said the case related to an earlier version of ChatGPT since enhanced with online search. noyb also told the BBC he had made several searches that day and that they could have influenced the answer; the full conversation was not published. No decision on the complaint had been reported by 1 November 2026.","incidentDate":"2024-08-01","incidentKind":"single_event","incidentDatePrecision":"month","exposurePattern":"single_interaction","reportedDate":"2025-03-20","aiSystem":"ChatGPT","aiProduct":"ChatGPT","aiCompany":"OpenAI","severity":"medium","verificationStatus":"verified","harmCategories":[],"harmOutcomes":["psychological_distress"],"harmOutcomeSummary":"A GDPR complaint says ChatGPT, asked who Arve Hjalmar Holmen is, produced a false account that he had murdered two of his sons; the complaint says he was deeply troubled, and he said the possibility that others would believe it scared him most.","frameworkFacets":[],"pathwayDirection":"inward","causationStatus":"alleged","participantUsersAffectedMin":1,"otherPeopleHarmedMin":0,"affectedCountStatus":"exact","affectedCountEvidence":"One harmed person: the complainant, who asked ChatGPT about himself and received the output. No other person is counted.","victimAgeRange":"adult","jurisdiction":"NO","platformType":"assistant","outcomeType":"regulatory_action","outcomeStatus":"ongoing","primarySourceUrl":"https://noyb.eu/en/ai-hallucinations-chatgpt-created-fake-child-murderer","primarySourceLabel":"noyb, 20 March 2025: AI hallucinations: ChatGPT created a fake child murderer","firstPublishedAt":"2026-01-13T10:13:04.624733+00:00","updatedAt":"2026-11-01T03:03:12.049479+00:00","scopeVersion":"facts-v3","tags":["chatgpt","openai","norway","gdpr","noyb","datatilsynet","false-claim","hallucination","regulator-complaint"]}]}