{ "meta": { "exportedAt": "2026-09-22T16:02:07.507Z", "formatVersion": 2, "selection": { "q": "police", "system": "", "harm": "", "context": "", "country": "", "role": "", "relation": "", "evidence": "", "year": "2024", "response": "", "severity": "", "verification": "", "view": "incidents", "sort": "added" }, "totalIncidents": 7, "coverage": { "cases": 7, "countries": 5, "languages": 3, "unknownLocation": 0, "locationPending": 0, "unknownLanguage": 0, "unknownDate": 0, "lawsuits": 2, "regulatory": 2, "minors": 2, "coreRelations": 3, "contextualRelations": 2, "mixedRelations": 0, "unknownRelations": 2, "relationPending": 1, "relationUnknown": 1 }, "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": "NOPE Incident Tracker", "url": "https://nope.net/incidents", "license": "CC BY 4.0" }, "incidents": [ { "id": "2026-netherlands-mrdeepfakes-74-year-old-prosecuted-deepfake-sex-videos-public-figures", "caseFacts": { "claims": [ { "id": "c1", "status": "corroborated", "evidence": [ { "locator": "'Managers en woordvoerders van verschillende slachtoffers bevestigen aan de NOS dat zij aangifte hebben gedaan of dat zij juridische stappen overwegen'", "relation": "supports", "source_id": "s4" }, { "locator": "'BBB-leider Caroline van der Plas, die bij Humberto zei dat de video voelt als \"digitale verkrachting\"'; '\"Ik doe geen aangifte voor mezelf, maar om nieuwe, jonge slachtoffers te voorkomen.\"'; 'Erachter komen dat ik in een deepfake pornovideo zit, was misselijkmakend.'", "relation": "supports", "source_id": "s5" }, { "locator": "'De zaak kwam in maart 2024 aan het licht. Vrouwelijke politici, olympische sporters, presentatrices en leden van het Koninklijk Huis doken op in gemanipuleerde pornovideo's'; 'Tientallen vrouwen deden aangifte.'", "relation": "supports", "source_id": "s1" } ], "assertion": "In March 2024 a large number of Dutch women in public life, including presenters and politicians, were found to appear in manipulated pornographic videos on an online platform; dozens filed complaints and several described the harm publicly.", "causal_attribution": "AD's 2024 investigation as reported by NOS, with victims' public statements." }, { "id": "c2", "status": "reported", "evidence": [ { "locator": "'Op zijn computer en andere gegevensdragers werden beelden aangetroffen van ongeveer zestig bekende Nederlanders en politici'; 'Naar aanleiding daarvan deden meer dan twintig mensen aangifte. De Noord-Hollander zou volgens het AD de meest actieve Nederlandse gebruiker van het platform zijn geweest'; 'Sommige video's werden tienduizenden keren bekeken'", "relation": "supports", "source_id": "s2" }, { "locator": "'De man kwam in beeld door onderzoek naar het beruchte platform MrDeepFakes. Dat leidde tot meer dan twintig aangiftes'", "relation": "supports", "source_id": "s3" } ], "assertion": "Images of about sixty well-known Dutch people and politicians were found on the suspect's computer and data carriers; AD's investigation into MrDeepFakes led to more than twenty complaints; according to AD he had been the platform's most active Dutch user, and some videos were viewed tens of thousands of times.", "causal_attribution": "AD's reporting of the OM's findings, relayed by NH Nieuws and Hart van Nederland; one chain." }, { "id": "c3", "status": "corroborated", "evidence": [ { "locator": "'Het Openbaar Ministerie gaat een 74-jarige man uit Noord-Holland vervolgen voor het maken van deepfakeporno van tientallen bekende Nederlandse vrouwen. Dat bevestigt een woordvoerder van het OM'; 'Na ruim een jaar onderzoek heeft justitie besloten hem voor de rechter te brengen'; 'De man riskeert een celstraf van maximaal twee jaar'", "relation": "supports", "source_id": "s1" }, { "locator": "'Het Openbaar Ministerie (OM) heeft na ruim een jaar onderzoek besloten de man voor de rechter te brengen'; 'De 74-jarige man kan daarvoor maximaal twee jaar gevangenisstraf krijgen'", "relation": "supports", "source_id": "s2" } ], "assertion": "After more than a year of investigation the OM decided to prosecute a 74-year-old man from Noord-Holland for making deepfake porn of dozens of well-known Dutch women; the maximum sentence is two years' imprisonment.", "causal_attribution": "OM spokesperson to NOS (after AD's report) and to AD (relayed by NH Nieuws); two chains for the decision itself." }, { "id": "c4", "status": "reported", "evidence": [ { "locator": "'Zij zegt tegen het AD opgelucht te zijn dat de zaak voor de rechter komt. Volgens haar heeft zij zelf ervaren welke impact dergelijke nepbeelden kunnen hebben.'", "relation": "supports", "source_id": "s2" } ], "assertion": "One of the presenters reported to be among the depicted said she was relieved the case would come to court, having experienced the impact of such fake images herself.", "causal_attribution": "Her statement to AD as relayed by NH Nieuws." }, { "id": "c5", "status": "reported", "evidence": [ { "locator": "'Meerdere vrouwelijke bekende Nederlanders overwegen juridische stappen omdat zij zijn opgedoken in gemanipuleerde pornovideo's. Deze video's staan op een online platform met maandelijks 13 miljoen bezoekers, schrijft het AD'; 'Nederlandse artiesten, tv-presentatrices, olympische sporters, (oud-)ministers, burgemeesters en leden van het Koninklijk Huis voorkomen'", "relation": "supports", "source_id": "s4" }, { "locator": "'De video's stonden op een online platform met maandelijks 13 miljoen bezoekers'", "relation": "context", "source_id": "s1" } ], "assertion": "AD reported that the platform had 13 million monthly visitors and that the depicted people included Dutch artists, television presenters, Olympic athletes, current and former ministers, mayors and members of the royal family.", "causal_attribution": "AD's 2024 investigation, relayed by NOS; one chain." }, { "id": "c6", "status": "reported", "evidence": [ { "locator": "'Hij is op de hoogte van het besluit, zegt het OM tegen de krant'; 'De website waar de video's op stonden, werd opgericht en gerund door een Canadees uit Toronto. Hij verwijderde eerder al beelden op verzoek van het OM en blokkeerde de website voor Nederlandse IP-adressen'; 'De Canadees wordt niet vervolgd. Wat hij deed, was in Canada niet strafbaar. En het Nederlandse Openbaar Ministerie gaat niet proberen om hem naar Nederland te halen. Dat zou te veel politiecapaciteit vergen, laat het OM weten aan het AD.'", "relation": "supports", "source_id": "s1" }, { "locator": "'De verdachte zit niet vast. Het is nog niet bekend wanneer de zaak inhoudelijk voor de rechter komt'", "relation": "supports", "source_id": "s2" } ], "assertion": "The suspect has been informed of the decision and is not in custody; no hearing date is known. The platform's founder-operator, a Canadian from Toronto who had removed images at the OM's request and blocked Dutch IP addresses before the site went offline, will not be prosecuted because his conduct was not punishable in Canada and the OM will not try to bring him to the Netherlands.", "causal_attribution": "OM statements to AD, relayed by NOS and NH Nieuws; one chain." } ], "effects": [ { "label": "dozens of women in Dutch public life depicted without consent in pornographic deepfake videos on a platform AD said had 13 million monthly visitors; complaints filed; public statements of harm", "claim_id": "c1", "direction": "negative" }, { "label": "a 74-year-old man to be prosecuted after more than a year of investigation, facing up to two years' imprisonment", "claim_id": "c3", "direction": "negative" } ], "sources": [ { "id": "s1", "url": "https://nos.nl/artikel/2631746-om-gaat-74-jarige-man-vervolgen-voor-maken-deepfakeporno-van-bn-ers", "kind": "news_report", "access": "read", "language": "nl", "translation_note": "Read live in Dutch on 2026-09-21 (HTTP 200; published 2026-09-20 12:26). An OM spokesperson confirmed the prosecution decision to NOS after AD's reporting; NOS adds its own 2024 reporting and the OM's position on the platform operator. Translated by the reviewer.", "independence_group": "nos-2026" }, { "id": "s2", "url": "https://www.nhnieuws.nl/nieuws/363601/om-vervolgt-74-jarige-man-uit-noord-holland-voor-deepfake-seksvideos", "kind": "news_report", "access": "read", "language": "nl", "translation_note": "Read live in Dutch on 2026-09-21 (NH Nieuws, published 2026-09-20 15:35 CEST (13:35 UTC)). Relays AD's report (which spoke to an OM spokesperson): devices, 'about sixty' people, more than twenty complaints, most active Dutch user, a presenter's reaction. AD itself answered 403. Translated by the reviewer.", "independence_group": "ad" }, { "id": "s3", "url": "https://www.hartvannederland.nl/tech/nieuws/artikelen/man-rechter-deepfake-seksvideos-bners", "kind": "news_report", "access": "read", "language": "nl", "translation_note": "Read live in Dutch on 2026-09-21 (Hart van Nederland, published 2026-09-20 11:43). Relays AD; grouped with s2. Translated by the reviewer.", "independence_group": "ad" }, { "id": "s4", "url": "https://nos.nl/artikel/2513371-vrouwelijke-bn-ers-overwegen-aangifte-vanwege-deepfake-pornovideo-s", "kind": "news_report", "access": "read", "language": "nl", "translation_note": "Read live in Dutch on 2026-09-21 (NOS, 19 March 2024). NOS's own 2024 report on the AD revelations, with confirmations from victims' managers and the minister's reaction. Translated by the reviewer.", "independence_group": "nos-2024" }, { "id": "s5", "url": "https://nos.nl/artikel/2513552-wel-of-niet-zeggen-dat-je-in-een-neppornovideo-zit-bn-ers-worstelen-ermee", "kind": "news_report", "access": "read", "language": "nl", "translation_note": "Read live in Dutch on 2026-09-21 (NOS, 20 March 2024). Own reporting with victims' public statements and a quote given to NOS. Translated by the reviewer.", "independence_group": "nos-2024" }, { "id": "s6", "url": "https://www.ad.nl/binnenland/gepensioneerde-man-74-wordt-vervolgd-voor-maken-deepfake-seksvideos-van-zestig-bners~aa272972/", "kind": "news_report", "access": "unavailable", "language": null, "translation_note": "Original AD report of 19 September 2026; the live page answered HTTP 403 (DPG paywall) and the Internet Archive CDX service was offline on 2026-09-21. Its contents are known only through s2, s3 and s1. The 19 September date rests on Google News feed metadata, not on the body.", "independence_group": "ad" } ], "version": 1, "ai_roles": [ "others_use" ], "contexts": [ "privacy", "justice", "everyday_life" ], "unknowns": [ "When the videos were made and over what period; the tools used.", "The exact number of depicted people and of complainants ('about sixty'; 'more than twenty').", "The charges' legal basis and whether distribution as well as creation is charged; no hearing date.", "The suspect's identity beyond age and province is not reported and is not sought here.", "The AD original could not be read; its details are known through relays.", "The range start (19 March 2024) is the date the videos' existence was publicly revealed and complaints began; the videos were made and posted at unreported earlier dates. The range end (19 September 2026) is the date the prosecution decision was reported; the OM's decision date is not stated." ], "geography": { "basis": "The suspect is from Noord-Holland and the videos were made at his home (NH Nieuws citing AD); the depicted people are Dutch public figures; the platform was operated from Toronto (NOS). The prosecution decision is by the Dutch OM; no hearing has been scheduled, so no court country is recorded.", "court_countries": [], "event_countries": [ "NL" ], "affected_person_countries": [ "NL" ] }, "publication": { "basis": "Published under the 2026-09-15 charter as an image-based abuse case affecting identifiable people, with harm described publicly by several of the depicted women and a prosecution decision confirmed by the Public Prosecution Service to NOS. Five Dutch bodies read (two NOS chains and the AD relay chain); the AD original is unavailable. Depicted people who spoke publicly are quoted in locators but not named in the record; the suspect is not named.", "reviewed_on": "2026-09-21" }, "ai_involvement": { "basis": "NOS, NH Nieuws and Hart van Nederland describe the videos as deepfakes in which the faces of known women were placed on pornographic footage, and the OM's prosecution, as confirmed to NOS, is for making deepfake porn (Hart van Nederland says he is suspected of making and distributing the videos); the specific tools are not identified in any inspected source.", "status": "reported" }, "person_relations": [ "depicted_or_impersonated" ] }, "name": "Netherlands: the Public Prosecution Service will prosecute a 74-year-old man from Noord-Holland for making deepfake sex videos of dozens of Dutch public figures, including presenters, politicians and a member of the royal family (images of about sixty people were found on his devices), after the 2024 revelations about the MrDeepFakes platform", "summary": "In March 2024 the newspaper AD revealed that a large number of Dutch women in public life, including artists, television presenters, Olympic athletes, current and former ministers, mayors and members of the royal family, appeared in manipulated pornographic videos on an online platform with 13 million monthly visitors, later identified as MrDeepFakes. Dozens of women filed police complaints and several spoke publicly: one presenter said she was preparing a complaint, a party leader said the video felt like 'digital rape', another presenter said she would file a complaint 'because tomorrow it could happen to young girls of sixteen', a presenter-entrepreneur did so 'to prevent new, young victims', and a member of parliament, told of a video by the House security service, went public rather than 'keep it small'. On 19 and 20 September 2026 AD, and the Public Prosecution Service (OM) confirming to NOS, reported that after more than a year of investigation the OM will prosecute a 74-year-old man from Noord-Holland, who is not in custody, for making the videos (Hart van Nederland, citing AD, says he is suspected of making and distributing them); images of about sixty well-known Dutch people and politicians were found on his devices, AD reported that he had been the platform's most active Dutch user, and he faces up to two years' imprisonment. The platform's Canadian operator, who removed material at the OM's request and blocked Dutch IP addresses before the site went offline, will not be prosecuted. One of the presenters reported to be among the depicted told AD she was relieved the case would go to court, having experienced the impact of such images herself. No hearing date has been set.", "incidentDate": "2024-03-19", "incidentEndDate": "2026-09-19", "incidentKind": "bounded_series", "incidentDatePrecision": "range", "exposurePattern": "unknown", "reportedDate": "2024-03-19", "aiSystem": "Face-swap deepfake video tools (not identified); videos published on the MrDeepFakes platform", "severity": "high", "verificationStatus": "credible", "harmCategories": [], "harmOutcomes": [ "exploitation_or_abuse", "psychological_distress", "reputational_harm" ], "harmOutcomeSummary": "Dozens of women in Dutch public life (images of about sixty people were found on the suspect's devices) were depicted without consent in pornographic deepfake videos, some viewed tens of thousands of times; those who spoke publicly described the experience as sickening and as 'digital rape', and one was advised by parliamentary security to stay silent (NOS 2024; NH Nieuws and Hart van Nederland citing AD, 2026). The 2026 prosecution decision is the consequence for the alleged maker.", "frameworkFacets": [], "causationStatus": "supported", "participantUsersAffectedMin": 0, "otherPeopleHarmedMin": 20, "affectedCountStatus": "documented_minimum", "affectedCountEvidence": "NH Nieuws and Hart van Nederland, citing AD, report that the MrDeepFakes investigation led to more than twenty complaints ('meer dan twintig aangiften'), a documented lower bound of 20 depicted complainants; images of about sixty well-known people were found on the suspect's devices, but 'ongeveer zestig' is approximate and not all are confirmed as depicted victims. Documented minimum 20.", "victimAgeRange": "adult", "jurisdiction": "NL", "platformType": "other", "outcomeType": "criminal_charges", "outcomeStatus": "ongoing", "primarySourceUrl": "https://nos.nl/artikel/2631746-om-gaat-74-jarige-man-vervolgen-voor-maken-deepfakeporno-van-bn-ers", "primarySourceLabel": "NOS, 20 September 2026: OM gaat 74-jarige man vervolgen voor maken deepfakeporno van BN'ers (OM spokesperson confirms after AD's report)", "firstPublishedAt": "2026-09-21T04:11:25.939553+00:00", "scopeVersion": "facts-v3", "tags": [ "deepfake", "ncii", "sexual-deepfake", "mrdeepfakes", "public-figures", "politicians", "royal-family", "prosecution", "netherlands", "noord-holland", "depicted" ] }, { "id": "2026-dillon-faces-wrongful-arrest-florida", "caseFacts": { "claims": [ { "id": "c1", "status": "corroborated", "evidence": [ { "locator": "'took pictures of the screen with his cell phone… low resolution… partially shadowed and off-axis'; '93 percent confidence'.", "relation": "supports", "source_id": "s1" }, { "locator": "Warrant based on the FR hit + 'a statement from a restaurant employee who picked his photo out of a lineup'; 'police called Dillon in 2023, he denied involvement citing skin-cancer scars'.", "relation": "supports", "source_id": "s3" } ], "assertion": "The warrant rested on a 93%-confidence FACES result obtained from low-resolution cell-phone photos of the restaurant's surveillance screen (partially shadowed, off-axis) plus a restaurant employee's photo-lineup identification; police had called Dillon in 2023 and he denied involvement, citing skin-cancer scars.", "causal_attribution": "The evidentiary basis is the complaint's account, carried by two chains; whether '93%' was system-generated or an investigator's characterization is unverified (the suit says the system 'spit out' his photo as a 'possible match')." }, { "id": "c2", "status": "corroborated", "evidence": [ { "locator": "'arrested Aug 2024 in Fort Myers'; 'more than two months later, prosecutors dropped the charges'; 'it took an entire year… to take down his mug shot and expunge the arrest'.", "relation": "supports", "source_id": "s1" }, { "locator": "Arrest bodycam 'outside his San Carlos Park home' via Gulf Coast affiliate; the Dillon interview ('I don't wish this on my worst enemy').", "relation": "supports", "source_id": "s2" }, { "locator": "'charges… dropped… in October 2024'; 1 night jail, bonded out pledging his truck's title.", "relation": "supports", "source_id": "s3" } ], "assertion": "Dillon was arrested at his San Carlos Park home in August 2024, spent one night in jail and bonded out pledging his truck's title; prosecutors dropped the charges in October 2024 after work-alibi evidence; the arrest took about a year to expunge, with his mugshot lingering online.", "causal_attribution": "The custody and expungement timeline is consistent across three chains." }, { "id": "c3", "status": "corroborated", "evidence": [ { "locator": "'suit in M.D. Fla. vs. city of Jacksonville Beach + officers/officials incl. Sheriff T.K. Waters'.", "relation": "supports", "source_id": "s1" }, { "locator": "Suit 'includes the Pinellas County Sheriff's Office, which oversees the use of the AI facial recognition technology across Florida'.", "relation": "supports", "source_id": "s2" }, { "locator": "ACLU attorney Nathan Wessler interview; defendants: Jacksonville Beach PD + Jacksonville and Pinellas County sheriffs' offices.", "relation": "supports", "source_id": "s3" } ], "assertion": "The ACLU filed a federal lawsuit in the week of 8 June 2026 in the Middle District of Florida against the city of Jacksonville Beach, Jacksonville Sheriff T.K. Waters and the Pinellas County Sheriff's Office, seeking damages and policy safeguards.", "causal_attribution": "The suit's existence and parties are carried by three chains; the exact filing date within the week of 8 June is not pinned in the inspected sources." }, { "id": "c4", "status": "documented", "evidence": [ { "locator": "PCSO response passages ('patently false'; 'never \"matches\"'; 'rests solely with any officer').", "relation": "supports", "source_id": "s2" } ], "assertion": "The Pinellas County Sheriff's Office responded that a training-failure claim 'is patently false', that 'facial recognition results are never \"matches\"', and that liability 'rests solely with any officer who did so'.", "causal_attribution": "The agency's position is quoted directly and preserved as contrary evidence." }, { "id": "c5", "status": "reported", "evidence": [ { "locator": "'one of at least 14 people arrested since 2019'.", "relation": "supports", "source_id": "s1" }, { "locator": "'one of 15 known people'.", "relation": "supports", "source_id": "s3" } ], "assertion": "The ACLU counts Dillon as at least the 14th known person wrongfully arrested since 2019 from police reliance on facial recognition (Reason, 11 June); a later WUSF attorney interview puts the count at 15 known.", "causal_attribution": "The advocacy count is attributed to the ACLU; the two figures are recorded as stated on their dates." } ], "effects": [ { "label": "wrongful felony arrest on a 93% facial-recognition hit, with a year-long expungement fight", "claim_id": "c2", "direction": "negative" } ], "sources": [ { "id": "s1", "url": "https://reason.com/2026/06/11/aclu-sues-after-facial-recognition-falsely-identifies-florida-man-as-a-child-abductor/", "kind": "news_report", "access": "read", "language": "en", "translation_note": "Read in English on 2026-09-15 (11 Jun 2026, Joe Lancaster).", "independence_group": "reason" }, { "id": "s2", "url": "https://abcnews.com/GMA/News/man-sues-law-enforcement-alleging-ai-facial-recognition/story?id=133810835", "kind": "news_report", "access": "read", "language": "en", "translation_note": "Read in English on 2026-09-15 (12 Jun 2026, Hipos & Yu).", "independence_group": "abc-news" }, { "id": "s3", "url": "https://www.wusf.org/courts-law/2026-06-18/ai-facial-recognition-tool-led-to-florida-man-wrongful-arrest-lawsuit", "kind": "news_report", "access": "read", "language": "en", "translation_note": "Read in English on 2026-09-15 (18 Jun 2026, Lyden/Peddie).", "independence_group": "wusf" } ], "version": 1, "ai_roles": [ "institutional_use" ], "contexts": [ "justice" ], "unknowns": [ "The exact filing date within the week of 8 June 2026 and the docket number are not pinned in the inspected sources.", "Whether the '93%' figure was system-generated or an investigator's characterization is unverified.", "The lineup procedure's details (who administered it and how) are only partly reported.", "The expungement's completion date is given only as 'about a year' after the arrest.", "The actual perpetrator of the 2023 luring attempt is not reported as identified." ], "geography": { "basis": "Event: the warrant (Jacksonville Beach) and the arrest (San Carlos Park / Fort Myers), Florida. Affected person: Dillon, Florida. Court: the felony charge (dropped) and the federal suit in the Middle District of Florida.", "court_countries": [ "US" ], "event_countries": [ "US" ], "affected_person_countries": [ "US" ] }, "publication": { "basis": "Published under the 2026-09-15 charter as an institutional automated-decision case (made_decision_about / made_claim_about): a facial-recognition hit treated as warrant-grade identification, documented by three independent chains including the ACLU's own attorney and the vendor agency's response. Dillon spoke publicly; the PCSO denial is preserved as contrary evidence.", "reviewed_on": "2026-09-15" }, "ai_involvement": { "basis": "The warrant rested on the FACES 93%-confidence result plus a lineup identification, per the complaint as carried by Reason and WUSF; the vendor (Pinellas County Sheriff's Office FACES) is confirmed by WUSF, and the PCSO's own response acknowledges its technology's role while rejecting liability.", "status": "supported" }, "person_relations": [ "made_decision_about", "made_claim_about" ] }, "name": "Florida: 93% facial-recognition hit led to Robert Dillon's arrest as an attempted child abductor; charges dropped, ACLU suit filed", "summary": "In August 2024, Robert Dillon, a 52-year-old self-employed commercial crabber from the Fort Myers area, was arrested at his San Carlos Park home on a third-degree felony warrant for the attempted luring of a child at a Jacksonville Beach McDonald's in November 2023 — a crime he had no part in. The warrant rested on a 93%-confidence result from FACES, the Face Analysis Comparison and Examination System maintained by the Pinellas County Sheriff's Office, run against low-resolution cell-phone photos of a surveillance screen, plus a restaurant employee's photo-lineup identification. Dillon spent a night in jail and bonded out against his truck's title; prosecutors dropped the charges in October 2024 after work-alibi evidence, but it took about a year to expunge the arrest and remove his mugshot. In June 2026 the ACLU filed a federal lawsuit (M.D. Fla.) against Jacksonville Beach, the Jacksonville sheriff and the Pinellas County Sheriff's Office, calling Dillon one of at least 14 known wrongful arrests from police reliance on facial recognition since 2019.", "incidentDate": "2024-08-01", "incidentEndDate": "2024-10-01", "incidentKind": "bounded_series", "incidentDatePrecision": "range", "exposurePattern": "single_interaction", "reportedDate": "2026-06-11", "aiSystem": "FACES (Face Analysis Comparison and Examination System), Pinellas County Sheriff's Office", "aiCompany": "Pinellas County Sheriff's Office (FACES centralized database)", "severity": "medium", "verificationStatus": "credible", "harmCategories": [], "harmOutcomes": [ "legal_harm", "reputational_harm", "psychological_distress" ], "harmOutcomeSummary": "Arrested at home on a child-abduction-class felony from a 93% facial-recognition hit on grainy screen photos; a night in jail, two months to clear the charge, and about a year to expunge the record and remove the mugshot.", "frameworkFacets": [], "causationStatus": "supported", "participantUsersAffectedMin": 0, "otherPeopleHarmedMin": 1, "affectedCountStatus": "exact", "affectedCountEvidence": "One documented harmed person: Dillon, the misidentified arrestee (three independent chains). The child in the underlying luring attempt is a victim of the actual perpetrator, not of the misidentification, and is not counted.", "victimAgeRange": "adult", "jurisdiction": "US", "platformType": "other", "outcomeType": "lawsuit_filed", "outcomeStatus": "ongoing", "primarySourceUrl": "https://reason.com/2026/06/11/aclu-sues-after-facial-recognition-falsely-identifies-florida-man-as-a-child-abductor/", "primarySourceLabel": "Reason (11 Jun 2026, Joe Lancaster) — complaint-based account: the screen photos, the 93% hit, the arrest, the expungement, the suit", "firstPublishedAt": "2026-09-15T10:27:54.71948+00:00", "scopeVersion": "facts-v3", "tags": [ "made-decision-about", "made-claim-about", "facial-recognition", "wrongful-arrest", "justice", "florida", "aclu" ] }, { "id": "2026-lorenzano-nunez-facial-recognition-wrongful-arrest", "caseFacts": { "claims": [ { "id": "c1", "status": "corroborated", "evidence": [ { "locator": "'investigators ran Rosado's old Arizona MVD photo through facial recognition databases operated by the Arizona Department of Public Safety and the FBI. They received 250 possible matches and zeroed in on Lorenzano Nunez'; the press-release/video passage (victim's son flown in, his handcuffs used).", "relation": "supports", "source_id": "s1" }, { "locator": "'Javier Lorenzano-Nunez was arrested in October 2024'; the son's account of the arrest period.", "relation": "supports", "source_id": "s2" } ], "assertion": "Lorenzano Nunez was arrested in October 2024 for the 1998 murder of Sarah Carr after investigators ran the old suspect's MVD photo through facial recognition databases (Arizona DPS and FBI), received 250 possible matches and zeroed in on him; Phoenix police publicized the arrest with a press release and a special video featuring the victim's son.", "causal_attribution": "The arrest and the FR role are carried by both chains; the lawsuit's characterization of FR as 'the key evidence' is attributed to the filing and court records." }, { "id": "c2", "status": "corroborated", "evidence": [ { "locator": "'who spent nearly a year in jail after being arrested for a 1998 Phoenix murder'; 'All charges were quietly dismissed less than a year later after forensic evidence, including DNA and fingerprints, excluded him, records show.'", "relation": "supports", "source_id": "s1" }, { "locator": "'charges dropped against him, without prejudice, in August 2025'; MCAO: 'additional evidence that put into question his guilt beyond a reasonable doubt'.", "relation": "supports", "source_id": "s2" } ], "assertion": "He spent nearly a year in jail; charges were dismissed without prejudice in August 2025 after forensic evidence, including DNA and fingerprints, excluded him, and the county attorney said additional evidence put his guilt beyond a reasonable doubt.", "causal_attribution": "The dismissal and its stated reason are official-account material carried by both chains; 'without prejudice' leaves refiling open." }, { "id": "c3", "status": "reported", "evidence": [ { "locator": "The 2007 Puerto Rico passage ('They did not act on it', per attorney Ortega) and the 2017-analysis passage ('The results excluded Lorenzano Nunez on 2 latent prints and were inconclusive on 2 others... seven years before').", "relation": "supports", "source_id": "s1" } ], "assertion": "The lawsuit alleges Phoenix police ignored a 2007 Puerto Rico Police lead reporting a man named Gilbert Noel Sanchez Rosado in custody with the same date of birth and social security number as the suspect, and that Phoenix's own 2017 fingerprint analysis had excluded Lorenzano Nunez on two latent prints (inconclusive on two) seven years before his arrest.", "causal_attribution": "These are the lawsuit's allegations carried by one chain from court records; the county attorney declined to comment. A court has not ruled on them." }, { "id": "c4", "status": "documented", "evidence": [ { "locator": "MCAO statement passages ('pointed toward...', 'valuable investigative tool'); the notice-of-claim critique ('merely identified... as a possible lead').", "relation": "supports", "source_id": "s2" } ], "assertion": "The Maricopa County Attorney's Office said the evidence police provided when the case was submitted 'pointed toward Javier Lorenzano-Nunez' and called facial recognition software a 'valuable investigative tool'; the notice of claim counters that facial recognition merely identified him as a possible lead.", "causal_attribution": "Both institutional positions are quoted directly; the dispute over the weight given to the FR match is the case's core and is preserved." }, { "id": "c5", "status": "reported", "evidence": [ { "locator": "The arrest-video passage (Miller flown in, his handcuffs used, interviewed for the city's special video).", "relation": "supports", "source_id": "s1" }, { "locator": "'It's not the outcome we wanted, but it's not over with'; 'I wholeheartedly believe in the investigators'.", "relation": "supports", "source_id": "s2" } ], "assertion": "The victim's son, Garrett Miller, himself a police officer, was flown in for the arrest and his handcuffs were used; after the dismissal he said the outcome was heart-wrenching but that he believes in the investigators and that 'it's not over with'.", "causal_attribution": "Contrary context preserved per the charter: the family's trust in the investigation is recorded alongside the wrongful-arrest account." } ], "effects": [ { "label": "nearly a year wrongfully jailed on a facial-recognition-driven arrest, publicized by police before forensic exclusion", "claim_id": "c2", "direction": "negative" } ], "sources": [ { "id": "s1", "url": "https://www.abc15.com/news/local-news/investigations/man-falsely-arrested-with-facial-recognition-for-cold-case-murder-sues-phoenix-pd-mcao", "kind": "news_report_court_records", "access": "read", "language": "en", "translation_note": "Read in English on 2026-09-15 (datePublished 2026-07-02T14:46-07:00; 23 paragraphs).", "independence_group": "abc15" }, { "id": "s2", "url": "https://www.fox10phoenix.com/news/sarah-jane-carrs-family-vows-keep-believing-justice-after-murder-charges-dismissed", "kind": "local_tv_news", "access": "read", "language": "en", "translation_note": "Read in English on 2026-09-15 (datePublished 2026-06-01T21:20-07:00; 20 paragraphs).", "independence_group": "fox10" } ], "version": 1, "ai_roles": [ "institutional_use" ], "contexts": [ "justice" ], "unknowns": [ "The specific facial-recognition vendor/system is not named in the inspected reporting.", "Details from ABC15's earlier 'About Face' investigation (including any grand-jury or foreign-custody particulars) were not re-verified in this pass and are not claimed.", "The federal suit's filing date and docket number are not stated in the inspected sources (reported as filed by 2 July 2026).", "Whether he was held continuously from October 2024 to August 2025 ('nearly a year in jail' per ABC15) is reported without a custody breakdown.", "The 1998 murder of Sarah Carr remains unsolved; the 2007 Puerto Rico lead's current status is unknown." ], "geography": { "basis": "Event: Phoenix, Arizona (arrest, detention, dismissal). Affected person: Lorenzano Nunez. Court: Maricopa County proceedings, then the federal suit; Puerto Rico's 2007 contact was a lead, not a proceeding.", "court_countries": [ "US" ], "event_countries": [ "US" ], "affected_person_countries": [ "US" ] }, "publication": { "basis": "Published under the 2026-09-15 charter as an institutional automated-decision case (made_decision_about / made_claim_about): a facial-recognition hit treated as key arrest evidence, with nearly a year of wrongful detention, documented through court-records reporting by two independent local chains. The county attorney's defense and the victim family's trust are preserved as contrary context.", "reviewed_on": "2026-09-15" }, "ai_involvement": { "basis": "The lawsuit and court records, per ABC15, make facial recognition 'the key evidence used to arrest'; the county attorney calls it a 'valuable investigative tool' while the notice of claim says it merely identified a possible lead. The specific vendor system is not named in the inspected reporting.", "status": "supported" }, "person_relations": [ "made_decision_about", "made_claim_about" ] }, "name": "Phoenix: facial recognition match put Javier Lorenzano Nunez in jail for nearly a year for a 1998 murder; charges dismissed, federal suit filed", "summary": "Javier Lorenzano Nunez was arrested in October 2024 for the 1998 murder of Sarah Carr after investigators ran the old suspect's MVD photo through facial recognition databases, got 250 possible matches and zeroed in on him — with Phoenix police publicizing the arrest in a press release and a special video. Charges were dismissed without prejudice in August 2025 after forensic evidence, including DNA and fingerprints, excluded him. His lawsuit alleges Phoenix police knew from a 2017 fingerprint analysis that his prints did not match, and ignored a 2007 Puerto Rico lead on the original suspect with the same name, date of birth and social security number. He spent nearly a year in jail. The federal suit against the Phoenix Police Department and the Maricopa County Attorney's Office alleges gross negligence, false arrest, false imprisonment and defamation.", "incidentDate": "2024-10-01", "incidentEndDate": "2025-08-01", "incidentKind": "bounded_series", "incidentDatePrecision": "range", "exposurePattern": "single_interaction", "reportedDate": "2026-06-01", "aiSystem": "Facial recognition systems operated by the Arizona Department of Public Safety and the FBI", "aiCompany": "Arizona DPS / FBI (systems not named in inspected reporting)", "severity": "high", "verificationStatus": "credible", "harmCategories": [], "harmOutcomes": [ "legal_harm", "loss_of_autonomy", "reputational_harm", "psychological_distress" ], "harmOutcomeSummary": "Nearly a year wrongfully jailed on a facial-recognition-driven cold-case arrest that police publicized with a video and the victim's son's handcuffs, ending only when DNA and fingerprint evidence excluded him; he now alleges false arrest, false imprisonment and defamation.", "frameworkFacets": [], "causationStatus": "supported", "participantUsersAffectedMin": 0, "otherPeopleHarmedMin": 1, "affectedCountStatus": "exact", "affectedCountEvidence": "One documented harmed person: Lorenzano Nunez, the subject of the wrongful arrest (court records and two news chains). The victim's family, whose hopes were raised and dashed, is recorded in the narrative but not counted as harmed persons under the counting rules.", "victimAgeRange": "adult", "jurisdiction": "US", "platformType": "other", "outcomeType": "lawsuit_filed", "outcomeStatus": "ongoing", "primarySourceUrl": "https://www.abc15.com/news/local-news/investigations/man-falsely-arrested-with-facial-recognition-for-cold-case-murder-sues-phoenix-pd-mcao", "primarySourceLabel": "ABC15 Arizona (2 Jul 2026) — the federal suit, court records, the 2017 fingerprint exclusion and the 2007 Puerto Rico lead", "firstPublishedAt": "2026-09-15T09:35:16.968902+00:00", "scopeVersion": "facts-v3", "tags": [ "made-decision-about", "made-claim-about", "facial-recognition", "wrongful-arrest", "justice", "arizona" ] }, { "id": "2024-kanpur-alleged-ai-voice-impersonation-fraud", "caseFacts": { "claims": [ { "id": "2024-kanpur-alleged-ai-voice-impersonation-fraud-claim-1", "status": "reported", "evidence": [ { "locator": "Incident narrative, payment paragraphs.", "relation": "supports", "source_id": "2024-kanpur-alleged-ai-voice-impersonation-fraud-source-1" } ], "assertion": "Reporting says a seller transferred INR 100,000, including INR 40,000 borrowed from nearby traders.", "causal_attribution": "Reported attribution remains unestablished independently." }, { "id": "2024-kanpur-alleged-ai-voice-impersonation-fraud-claim-2", "status": "reported", "evidence": [ { "locator": "Police-investigation paragraph and closing police quotation.", "relation": "supports", "source_id": "2024-kanpur-alleged-ai-voice-impersonation-fraud-source-1" } ], "assertion": "Reporting says police registered a case and began investigating.", "causal_attribution": "Reported proceeding; this does not establish AI causation." } ], "effects": [ { "label": "money_transferred", "claim_id": "2024-kanpur-alleged-ai-voice-impersonation-fraud-claim-1", "direction": "negative" } ], "sources": [ { "id": "2024-kanpur-alleged-ai-voice-impersonation-fraud-source-1", "url": "https://www.amarujala.com/uttar-pradesh/kanpur/cheated-for-the-first-time-by-deepfake-ai-in-kanpur-told-uncle-in-nephews-voice-save-me-by-any-means-2024-04-10", "kind": "local_reporting", "access": "read", "language": "hi", "translation_note": "Model read the original language. No independent bilingual human review.", "independence_group": "2024-kanpur-alleged-ai-voice-impersonation-fraud-account" } ], "version": 1, "ai_roles": [ "others_use" ], "contexts": [ "finance" ], "unknowns": [ "No recording, forensic report, bank proof or police record inspected.", "Final loss after any recovery is unknown.", "The caller’s allegation about a relative is a scam claim, not adopted as a true offence.", "The event year follows the dated article context; the narrative explicitly gives 27 March." ], "geography": { "basis": "Narrative explicitly locates the seller and episode in Kanpur, India.", "court_countries": [], "event_countries": [ "IN" ], "affected_person_countries": [ "IN" ] }, "publication": { "basis": "A concrete adverse account with an inspected source, explicit attribution and uncertainty. Duplicate and identity review completed.", "reviewed_on": "2026-09-15" }, "ai_involvement": { "basis": "Closing police quotation gives only a preliminary AI attribution.", "status": "suspected" }, "person_relations": [ "unknown" ] }, "name": "Kanpur seller reportedly transfers money after alleged AI voice impersonation", "summary": "Hindi reporting describes a seller transferring INR 100,000 after an impersonation call. A quoted police assessment treats AI involvement as preliminary. The transfers and AI attribution remain independently unverified.", "incidentDate": "2024-03-27", "incidentKind": "single_event", "incidentDatePrecision": "day", "exposurePattern": "unknown", "reportedDate": "2024-04-10", "aiSystem": "Unspecified voice-generation system", "severity": "medium", "verificationStatus": "unverified", "harmCategories": [], "harmOutcomes": [ "financial_loss" ], "harmOutcomeSummary": "A seller reportedly transferred INR 100,000.", "frameworkFacets": [], "causationStatus": "alleged", "participantUsersAffectedMin": 0, "otherPeopleHarmedMin": 1, "affectedCountStatus": "exact", "affectedCountEvidence": "One seller is described as harmed; other people are not counted.", "victimAgeRange": "unknown", "outcomeType": "investigation_opened", "outcomeStatus": "unknown", "primarySourceUrl": "https://www.amarujala.com/uttar-pradesh/kanpur/cheated-for-the-first-time-by-deepfake-ai-in-kanpur-told-uncle-in-nephews-voice-save-me-by-any-means-2024-04-10", "primarySourceLabel": "Amar Ujala, Hindi reporting", "firstPublishedAt": "2026-09-15T03:07:27.856477+00:00", "scopeVersion": "facts-v3", "tags": [ "finance", "impersonation" ] }, { "id": "2024-aligarh-alleged-voice-cloning-fraud", "caseFacts": { "claims": [ { "id": "2026-09-15-hi-voice-fraud-claim-1", "status": "reported", "evidence": [ { "locator": "Article body, incident narrative and closing police statement", "relation": "supports", "source_id": "2026-09-15-hi-voice-fraud-source" } ], "assertion": "Local reporting says a parent transferred INR 400,000 after a call claiming his son was in police trouble. The report says police stopped one INR 86,000 transaction.", "causal_attribution": "Attributed in the account; independent causation has not been established." } ], "effects": [ { "label": "money_transferred", "claim_id": "2026-09-15-hi-voice-fraud-claim-1", "direction": "negative" } ], "sources": [ { "id": "2026-09-15-hi-voice-fraud-source", "url": "https://www.amarujala.com/uttar-pradesh/aligarh/cheated-four-lakhs-by-creating-son-voice-with-ai-2024-02-01", "kind": "local_reporting", "access": "read", "language": "hi", "translation_note": "Model read the original language. No separate bilingual human review.", "independence_group": "2026-09-15-hi-voice-fraud-account" } ], "version": 1, "ai_roles": [ "others_use" ], "contexts": [ "finance", "fraud" ], "unknowns": [ "AI technique is attributed to a police statement in reporting; no recording or forensic evidence read.", "Final net loss and any later recovery are unknown.", "The son's alleged offence was part of the scam narrative, not an allegation adopted by this dossier." ], "geography": { "basis": "Report explicitly places the parent and incident in Aligarh, India.", "court_countries": [], "event_countries": [ "IN" ], "affected_person_countries": [ "IN" ] }, "publication": { "basis": "Source body re-read, event boundary and duplicates checked, identity details minimized, and claim uncertainty stated.", "reviewed_on": "2026-09-15" }, "ai_involvement": { "basis": "AI use is attributed in the source; no independent technical verification performed.", "status": "reported" } }, "name": "Aligarh parent reportedly transfers money after an alleged AI voice-impersonation call", "summary": "Hindi reporting describes a parent in Aligarh transferring INR 400,000 after callers impersonated police and used a voice presented as his son’s. Police reportedly stopped INR 86,000. The AI attribution and transfers have not been independently verified here; final net loss is unknown.", "incidentDate": "2024-01-30", "incidentKind": "single_event", "incidentDatePrecision": "day", "exposurePattern": "unknown", "reportedDate": "2024-02-02", "aiSystem": "Unspecified voice-generation system", "severity": "medium", "verificationStatus": "unverified", "harmCategories": [], "harmOutcomes": [ "financial_loss" ], "harmOutcomeSummary": "Local reporting says a parent transferred INR 400,000 after a call claiming his son was in police trouble. The report says police stopped one INR 86,000 transaction.", "frameworkFacets": [], "causationStatus": "alleged", "participantUsersAffectedMin": 0, "otherPeopleHarmedMin": 1, "affectedCountStatus": "exact", "affectedCountEvidence": "One person is described as harmed in this account; wider affected populations are not counted.", "victimAgeRange": "unknown", "outcomeStatus": "unknown", "primarySourceUrl": "https://www.amarujala.com/uttar-pradesh/aligarh/cheated-four-lakhs-by-creating-son-voice-with-ai-2024-02-01", "primarySourceLabel": "Amar Ujala, Hindi reporting", "firstPublishedAt": "2026-09-15T01:18:49.295668+00:00", "scopeVersion": "facts-v3", "tags": [ "finance", "fraud" ] }, { "id": "2025-valencia-spain-deepfake-school", "caseFacts": { "claims": [ { "id": "c1", "status": "reported", "evidence": [ { "locator": "'began investigating in December last year after a female student reported the creation of a social media account in her name that featured an AI-generated video'; '\"All these photos had been modified from the originals, which had been manipulated so that the people in them appeared completely naked\"'; 'a webpage devoted to advertising the images for commercial purposes'; 'After 15 other minors came forward with similar complaints'; 'a 17-year-old boy was being investigated on suspicion of corrupting minors'.", "relation": "supports", "source_id": "s1" } ], "assertion": "From December 2024 a girl in the Ribera Alta area reported fake accounts in her name with an AI-generated video and AI-manipulated nude photos of minors, plus a webpage advertising the images commercially; 15 other minors complained; the Guardia Civil traced the accounts to a 17-year-old schoolmate, now investigated for corrupting minors under a Valencia youth court.", "causal_attribution": "Guardia Civil statement as reported by the Guardian; single chain." } ], "effects": [ { "label": "16 minors depicted in AI-generated nude images and videos on fake accounts, one impersonated by name, images advertised for sale", "claim_id": "c1", "direction": "negative" } ], "sources": [ { "id": "s1", "url": "https://www.theguardian.com/world/2025/jul/27/spanish-teenager-investigated-ai-generated-nude-videos", "kind": "news_report", "access": "read", "language": "en", "translation_note": "Read in English on 2026-09-18 (The Guardian, 27 July 2025). Reports and quotes the Guardia Civil's statement of 27 July 2025; the Spanish original of the statement was not read.", "independence_group": "guardian" } ], "version": 1, "ai_roles": [ "others_use" ], "contexts": [ "education", "privacy", "justice" ], "unknowns": [ "The outcome of the youth-court proceeding is not known.", "The AI tools and the platforms hosting the accounts are not named.", "Whether any images were actually sold is not stated." ], "geography": { "basis": "Guardia Civil officers in the Ribera Alta area of Valencia province investigated; the victims are the suspect's schoolmates there; a youth court in Valencia oversees the case (Guardian).", "court_countries": [ "ES" ], "event_countries": [ "ES" ], "affected_person_countries": [ "ES" ] }, "publication": { "basis": "Restored under the 2026-09-15 charter as a harmful-artifact case involving identified (unnamed) depicted minors, an impersonation and a police investigation under a youth court. One chain read; claims recorded as reported. Original ID and first-publication timestamp preserved; incident date corrected to the first complaint.", "reviewed_on": "2026-09-18" }, "ai_involvement": { "basis": "The Guardia Civil statement, as quoted by the Guardian, describes AI-generated videos and AI-modified photos and a website advertising AI image-modification services; no forensic detail is published. The artifacts depicted the minors and impersonated one girl by name.", "status": "reported" }, "person_relations": [ "depicted_or_impersonated" ] }, "name": "Ribera Alta, Valencia: a 17-year-old was investigated for making AI-generated nude images and videos of 16 schoolmates, impersonating a girl on fake accounts and advertising the images for sale", "summary": "In December 2024 a female student in the Ribera Alta area of Valencia province reported to the Guardia Civil that a social-media account had been created in her name featuring an AI-generated video, and later a second account in her name, other accounts showing AI-generated nude videos, and a webpage advertising the images for commercial purposes; the Guardia Civil said the photos, all of minors, had been manipulated so that the people in them appeared completely naked. After 15 other minors came forward with similar complaints, officers traced IP addresses and account emails to the home of a 17-year-old schoolmate, who is being investigated on suspicion of corrupting minors under a Valencia youth court. The Guardia Civil announced the investigation on 27 July 2025.", "incidentDate": "2024-12-01", "incidentKind": "bounded_series", "incidentDatePrecision": "month", "exposurePattern": "repeated_interactions", "reportedDate": "2025-07-27", "aiSystem": "AI image and video manipulation tools (not named; police examined a website advertising AI image-modification services)", "severity": "high", "verificationStatus": "credible", "harmCategories": [], "harmOutcomes": [ "exploitation_or_abuse", "psychological_distress", "reputational_harm" ], "harmOutcomeSummary": "Sixteen minors were depicted in AI-generated nude images and videos published on fake accounts, with one girl also impersonated by name, and the images were advertised for sale; a schoolmate is under youth-court investigation.", "frameworkFacets": [], "causationStatus": "supported", "participantUsersAffectedMin": 0, "otherPeopleHarmedMin": 16, "affectedCountStatus": "exact", "affectedCountEvidence": "One complainant plus 15 other minors who came forward with similar complaints (Guardian, quoting the Guardia Civil): exact count of 16 depicted minors.", "victimAgeRange": "minor", "jurisdiction": "ES", "platformType": "other", "outcomeType": "investigation_opened", "outcomeStatus": "ongoing", "primarySourceUrl": "https://www.theguardian.com/world/2025/jul/27/spanish-teenager-investigated-ai-generated-nude-videos", "primarySourceLabel": "The Guardian, 27 July 2025: Spanish teenager investigated over AI-generated nude videos of 16 schoolmates (Guardia Civil statement)", "firstPublishedAt": "2026-03-09T05:45:25.741735+00:00", "scopeVersion": "facts-v3", "tags": [ "deepfake", "ai-generated-imagery", "minors", "school", "impersonation", "spain", "valencia", "guardia-civil", "youth-court", "depicted" ] }, { "id": "2024-south-korea-telegram-deepfake-crisis", "caseFacts": { "claims": [ { "id": "c1a", "status": "corroborated", "evidence": [ { "locator": "BBC 3 Sep 2024 confirms Ko Narin's Hankyoreh scoop and describes the school- and university-specific rooms.", "relation": "supports", "source_id": "s1" }, { "locator": "Hankyoreh English's own exposure of the military Telegram room (27 Aug 2024).", "relation": "supports", "source_id": "s3" } ], "assertion": "Around 26 August 2024 The Hankyoreh (Ko Narin) exposed Telegram rooms creating and sharing AI-generated sexual deepfakes of Korean women and girls, with rooms organized by school and university.", "causal_attribution": "The scoop's existence and the rooms' school/university organization are confirmed by the BBC's original reporting and Hankyoreh English's companion investigation; the original Korean article is dead-linked and retained uninspected." }, { "id": "c1b", "status": "reported", "evidence": [ { "locator": "Room-structure passages: 'humiliation rooms'/'friend of friend rooms', entry terms, the 2,000-member school group.", "relation": "supports", "source_id": "s1" } ], "assertion": "The rooms ran as 'humiliation rooms' and 'friend of friend rooms' with entry terms demanding at least four photos plus name, age and area, and one group for underage pupils at a single school had more than 2,000 members.", "causal_attribution": "Single-chain detail from the BBC's 3 Sep 2024 account; no second inspected chain describes the entry terms or the 2,000-member group." }, { "id": "c2", "status": "reported", "evidence": [ { "locator": "Police-figure passages: 297 vs 180; 'The victims are often minors and the perpetrators are mostly teenagers'.", "relation": "supports", "source_id": "s2" }, { "locator": "Watchdog ~6,500 requests Jan-Jul (4x); 178 suspects booked, 74% aged 10-19; >half of erased deepfakes involved minors.", "relation": "supports", "source_id": "s4" } ], "assertion": "Police recorded 297 deepfake cases in January-July 2024 against 180 in all of 2023; the watchdog received nearly 6,500 takedown requests in the first seven months of 2024 (four times the prior year); of 178 suspects booked, 74% were aged 10-19; more than half of erased deepfakes involved minors; victims were often minors and perpetrators mostly teenagers.", "causal_attribution": "Both outlets relay the same police and watchdog accounts; traced to one official origin, so the figures are reported rather than corroborated. Figures count cases, requests and suspects, not harmed people." }, { "id": "c3", "status": "reported", "evidence": [ { "locator": "'more than 900 students, teachers and staff' (education ministry); 'of 964 deepfake-related sex crime cases reported from January to October last year, police made 23 arrests' (Seoul police statement).", "relation": "supports", "source_id": "s6" } ], "assertion": "The education ministry counted more than 900 students, teachers and staff victimized between January and early November 2024; a Seoul police statement said that of 964 deepfake-related sex crime cases reported January-October 2024, 23 arrests were made.", "causal_attribution": "Single chain (CNN) relaying official figures; the education-ministry count is the record's person-count lower bound." }, { "id": "c4", "status": "corroborated", "evidence": [ { "locator": "Reuters wire via CNN, 26 Sep 2024: possession/viewing criminalized.", "relation": "supports", "source_id": "s5" }, { "locator": "CNN 25 Apr 2025: Sep 2024 law — possession/viewing up to 3 years or 30M won; creation maximum raised 5 to 7 years.", "relation": "supports", "source_id": "s6" } ], "assertion": "On 26 September 2024 the National Assembly criminalized possessing or viewing sexually explicit deepfakes (up to 3 years' prison or a 30 million won fine); the maximum sentence for creating deepfake pornography was raised from 5 to 7 years.", "causal_attribution": "The statute is a public record carried by two chains (Reuters wire; CNN original). The historical claim of an October 2024 amendment removing the distribution-intent requirement was not re-verified and is recorded as an unknown." }, { "id": "c5", "status": "reported", "evidence": [ { "locator": "'Military IDs turned into deepfake porn': servicewomen called 'military supplies'; at least 30 victims; channel active since 8 Aug with ~850 audience.", "relation": "supports", "source_id": "s3" } ], "assertion": "A separate Telegram room degraded servicewomen with deepfakes made from military-intranet ID photos, implying insider involvement, with at least 30 victims across the Army, Navy and Air Force and an audience of about 850 since 8 August 2024.", "causal_attribution": "Single chain (Hankyoreh original reporting)." }, { "id": "c6", "status": "reported", "evidence": [ { "locator": "Heejin passages (pseudonymous victim account).", "relation": "supports", "source_id": "s1" }, { "locator": "Ruma passages (pseudonymous victim account).", "relation": "supports", "source_id": "s6" } ], "assertion": "Individual victims described severe harm: a pseudonymous school victim 'Heejin' spoke to the BBC, and 'Ruma', targeted in 2021 at age 27, spoke to CNN.", "causal_attribution": "Pseudonymous first-person accounts carried by two outlets; the historical record's 'broke my whole belief system' quotation was not re-found in inspected sources and is dropped from the claims." }, { "id": "c7", "status": "reported", "evidence": [ { "locator": "Korea Herald 16 Nov 2025 relaying the National Office of Investigation figures.", "relation": "supports", "source_id": "s7" } ], "assertion": "By November 2025 the National Office of Investigation reported 3,557 people apprehended for cybersexual violence over the preceding year, with deepfake crimes the largest category (1,553 cases) and about 62% of suspects teenagers.", "causal_attribution": "Single chain relaying official figures; apprehension figures count suspects, not victims." } ], "effects": [ { "label": "mass AI-generated sexual deepfakes of identified women and girls distributed in Telegram rooms", "claim_id": "c1a", "direction": "negative" } ], "sources": [ { "id": "s1", "url": "https://www.bbc.com/news/articles/cpdlpj9zn9go", "kind": "news_report", "access": "read", "language": "en", "translation_note": "Read in English on 2026-09-15 (3 Sep 2024; Jean Mackenzie, Seoul). Original reporting: victim 'Heejin', room structure, the 2,000-member school group, the 500+ activist-compiled school list.", "independence_group": "bbc" }, { "id": "s2", "url": "https://www.bbc.com/news/articles/cg4yerrg451o", "kind": "news_report", "access": "read", "language": "en", "translation_note": "Read in English on 2026-09-15 (28 Aug 2024). Yoon cabinet order; police figures attributed.", "independence_group": "bbc" }, { "id": "s3", "url": "https://english.hani.co.kr/arti/english_edition/e_national/1155588.html", "kind": "news_report", "access": "read", "language": "en", "translation_note": "Read in English on 2026-09-15 (27 Aug 2024). Hankyoreh's own exposure of the military room.", "independence_group": "hankyoreh" }, { "id": "s4", "url": "https://www.npr.org/2024/09/06/nx-s1-5101891/south-korea-deepfake", "kind": "news_report", "access": "read", "language": "en", "translation_note": "Read in English on 2026-09-15 (6 Sep 2024). Watchdog and police figures attributed.", "independence_group": "npr" }, { "id": "s5", "url": "https://www.cnn.com/2024/09/26/asia/south-korea-deepfake-bill-passed-intl-hnk", "kind": "wire_report", "access": "read", "language": "en", "translation_note": "Read in English on 2026-09-15 (26 Sep 2024); bylined Reuters.", "independence_group": "reuters" }, { "id": "s6", "url": "https://www.cnn.com/2025/04/25/asia/south-korea-deepfake-crimes-intl-hnk-dst", "kind": "news_report", "access": "read", "language": "en", "translation_note": "Read in English on 2026-09-15 (25 Apr 2025). CNN original: victim 'Ruma', education-ministry figure, Seoul police statement.", "independence_group": "cnn" }, { "id": "s7", "url": "https://www.koreaherald.com/article/10616925", "kind": "news_report", "access": "read", "language": "en", "translation_note": "Read in English on 2026-09-15 (16 Nov 2025; the historical date 2025-11-01 was wrong). National Office of Investigation figures.", "independence_group": "korea-herald" }, { "id": "s8", "url": "https://www.hani.co.kr/arti/society/society_general/1155800.html", "kind": "news_report", "access": "unavailable", "language": null, "translation_note": "HTTP 404 on 2026-09-15. Ko Narin's original Korean investigation (~26 Aug 2024); retained as the historical reference. BBC's 3 Sep 2024 report confirms the scoop; the body was not inspected in this review and its language is not re-asserted.", "independence_group": "hankyoreh" } ], "version": 1, "ai_roles": [ "others_use" ], "contexts": [ "education", "justice" ], "unknowns": [ "Not re-verified in inspected sources and therefore not claims of this record: the 220,000-member channel figure; 781 victims with 37% minors; Women's Human Rights Institute of Korea 332,341 supported cases; the October 2024 amendment removing the distribution-intent requirement; the Seoul National University graduates' 9-year and 3.5-year sentences; the teachers' union 200-school estimate; the 'broke my whole belief system' victim quotation.", "The specific AI tools used are not identified in the inspected reporting.", "Ko Narin's original Korean article (hani.co.kr 1155800) is HTTP 404; its contents are known only through the BBC's confirmation and the historical record.", "The ScienceDirect study cited by the historical record answered 403 and was not inspected.", "Cross-group overlap between police case counts, the education-ministry victim count and watchdog requests is unresolved." ], "geography": { "basis": "Events, victims and the responding institutions are all in South Korea (all chains). No court proceeding is verified in this review, so court_countries is [].", "court_countries": [], "event_countries": [ "KR" ], "affected_person_countries": [ "KR" ] }, "publication": { "basis": "Restored under the 2026-09-15 charter as a harmful-artifact crisis affecting identified women and girls, predominantly minors. The person count is the education ministry's documented lower bound; case, request and membership figures are recorded but never used as person counts. Unverifiable historical claims are preserved as unknowns rather than deleted silently. No death; the death-verification rule is not engaged. Pseudonymous victims remain pseudonymous. In a 2026-09-15 second pass the original compound first claim was split: the corroborated exposure and room organization stand as c1a, and the BBC-only structure details (entry terms, 2,000-member group) stand as reported c1b. person_relations recorded in the same pass.", "reviewed_on": "2026-09-15" }, "ai_involvement": { "basis": "Every inspected chain describes AI deepfake tools as the generation mechanism and Telegram as the distribution channel; police and watchdog figures classify the cases as deepfake sex crimes. The specific tools are not individually identified beyond 'AI deepfake/nudify tools'.", "status": "supported" }, "person_relations": [ "depicted_or_impersonated" ] }, "name": "South Korea Telegram AI Deepfake Sexual Abuse Crisis", "summary": "In late August 2024, journalist Ko Narin of The Hankyoreh exposed networks of Telegram rooms in South Korea where AI tools were used to turn ordinary photos of women and girls — middle and high school pupils, university students, teachers and servicewomen — into sexual deepfakes, shared with their identities in school- and university-specific 'humiliation rooms'. Police figures showed 297 deepfake cases in January-July 2024 (vs 180 in all of 2023); the education ministry counted more than 900 victimized students, teachers and staff; President Yoon ordered a crackdown; and on 26 September 2024 the National Assembly criminalized possessing or viewing sexually explicit deepfakes. Victims were often minors and perpetrators mostly teenagers.", "incidentDate": "2024-08-01", "incidentEndDate": "2024-09-26", "incidentKind": "bounded_series", "incidentDatePrecision": "range", "exposurePattern": "repeated_interactions", "reportedDate": "2024-08-26", "aiSystem": "AI deepfake generation tools (various)", "aiCompany": "Multiple (various free AI 'undressing' tools)", "severity": "critical", "verificationStatus": "credible", "harmCategories": [ "minor_exploitation", "third_party_harm_facilitation", "psychological_manipulation" ], "harmOutcomes": [ "psychological_distress", "reputational_harm" ], "harmOutcomeSummary": "Mass non-consensual AI sexual imagery of schoolgirls, students, teachers and servicewomen, distributed with their identities in Telegram rooms; pseudonymous victims describe severe distress, and police recorded hundreds of cases dominated by minor victims and teenage perpetrators.", "frameworkFacets": [], "causationStatus": "supported", "participantUsersAffectedMin": 0, "otherPeopleHarmedMin": 900, "affectedCountStatus": "documented_minimum", "affectedCountEvidence": "Lower bound: the South Korean education ministry's count of 'more than 900 students, teachers and staff' victimized (January-early November 2024), carried by CNN's 25 April 2025 report. This counts people reported as victims to the education authorities. Police figures (297 cases; 964 cases) count cases, not people, and are not used as person counts; the watchdog's ~6,500 figures count takedown requests, not people; the 500+ schools figure is an activist-compiled list of targets, not a victim count; channel membership numbers are not counted. Cross-group overlap is unresolved.", "victimAgeRange": "minor", "jurisdiction": "KR", "platformType": "other", "outcomeType": "criminal_charges", "outcomeStatus": "ongoing", "primarySourceUrl": "https://www.bbc.com/news/articles/cpdlpj9zn9go", "primarySourceLabel": "BBC News, Jean Mackenzie (3 Sep 2024) — 'Inside the deepfake porn crisis engulfing Korean schools' (replaces the dead Hankyoreh original)", "firstPublishedAt": "2026-01-29T09:52:18.956616+00:00", "scopeVersion": "facts-v3", "tags": [ "deepfakes", "minor", "south-korea", "telegram", "sexual-images", "school", "criminal-charges", "legislation", "mass-scale", "non-western" ] } ] }