⚠️ AI Incidents
Documented real-world AI failures and harms — the cases that show why evals, guardrails and policy exist. Live from the AI Incident Database, refreshed hourly.
- H.R. v. Allegheny Health Network and Highmark Health (Class Action Complaint)Nov 18, 2026
AIID editor's note (AI transparency): Abridged from the supplied filing using OpenAI GPT-6.2 Sol (gpt-6.2-sol). The excerpt retains the allegations about the patient's visits, the Abridge deployment, the consent process and the claimed harm ... (https://incidentdatabase.ai/cite/1
- Allegheny Health Network, Highmark Health Used AI Documentation Tool Without Consent, Class Action Suit ClaimsNov 18, 2026
A proposed class action lawsuit alleges that Allegheny Health Network and parent company Highmark Health have violated a Pennsylvania privacy law by using an AI-powered ambient recording tool to document conversations between patients and t ... (https://incidentdatabase.ai/cite/1
- 123 Reg discloses 'AI cyberattack' on customer data, suspects China nexusNov 18, 2026
UK domain registrar 123 Reg (no relation – Ed) says it was the victim of an “AI-driven cyberattack” that exposed customer data. Customers began receiving their notification emails this week, informing them that their contact information, " ... (https://incidentdatabase.ai/cite/1
- GoDaddy Says Its Been Penetrated by its First AI-Driven Cyber AttackNov 18, 2026
TEMPE, Ariz.–GoDaddy, which serves as a web host and provides other services to credit unions, is reporting that on Sept. 18 it experienced an AI-driven cyberattack that resulted in exposure of information about users’ Websites + Marketing ... (https://incidentdatabase.ai/cite/1
- GoDaddy AI cyberattack 2026, what was exposed and what website owners should do nowNov 18, 2026
The GoDaddy AI cyberattack of 18 September 2026 did not take passwords, but it did hand an attacker a structured list of GoDaddy website builder customers and the contact details those customers publish. That combination is what matters. Ac ... (https://incidentdatabase.ai/cite/1
- Maricopa County court releases DUI suspect after AI-generated police reportNov 16, 2026
PHOENIX — A Maricopa County judge released a DUI suspect after determining he could not find probable cause in the case because the arresting officer's report was generated with the assistance of an artificial intelligence tool. Anthony Ed ... (https://incidentdatabase.ai/cite/1
- DUI suspect released after officer used AI in arrest report: JudgeNov 16, 2026
An Arizona judge released a man who was accused of driving under the influence after it was revealed that the officer who arrested the man used artificial intelligence in his arrest report. Maricopa Superior Court Judge Daniel Fenzel order ... (https://incidentdatabase.ai/cite/1
- Maricopa County judge dismisses DUI case after trooper uses AINov 16, 2026
PHOENIX (AZFamily) — A Maricopa County judge dismissed a DUI case against Anthony Quidera and explained that the trooper involved used AI to write the probable cause statement. Quidera was released as a result. More police departments acro ... (https://incidentdatabase.ai/cite/1
- Ioan Gruffudd Says ‘Explicit and Disturbing’ AI Scam Convinced Fan They Had an Illicit RomanceNov 16, 2026
Ioan Gruffudd warned his fans online of an AI scammer using his likeness after one woman believed she was in an “extramarital relationship” with the actor after hundreds of telegram messages and AI-generated images were shared between the t ... (https://incidentdatabase.ai/cite/1
- Ioan Gruffudd warns of ‘distressing’ scams using ‘explicit’ AI content of himNov 16, 2026
Welsh actor Ioan Gruffudd has warned fans about “distressing” scams involving “explicit and disturbing” AI-generated images and videos of his likeness. The 53-year-old, known for playing Reed Richards in the Fantastic Four film series, has ... (https://incidentdatabase.ai/cite/1
- Ioan Gruffudd Calls for A.I. Regulation After Fan Gets Scammed By ImposterNov 16, 2026
Ioan Gruffudd is calling for Artificial Intelligence regulation after an uncomfortable encounter with a fan who'd fallen for an A.I. scam and believed they'd been having an affair. We caught up with the actor at LAX on Monday, where he gav ... (https://incidentdatabase.ai/cite/1
- Passengers injured after collision between bus and delivery robot in TallinnNov 16, 2026
The incident took place at around 6 p.m. at a pedestrian crossing at the intersection of Luise and Tehnika streets in the city center, municipal transport company Tallinn Transport (TLT) said. The Bolt delivery robot approached from Tehnik ... (https://incidentdatabase.ai/cite/1
- Starship plans software changes to improve delivery robots' safetyNov 16, 2026
A bus and a delivery robot collided in Tallinn on Monday evening, injuring several people after the driver braked sharply to avoid an accident. Tallinn Transport (TLT) board member Andrei Novikov said the accident occurred during rush hour ... (https://incidentdatabase.ai/cite/1
- TV3 uudised: avarii pakiroboti ja bussi vahel. Uurimine tõestas, et pakirobot ei põgenenud sündmuskohaltNov 16, 2026
Tallinnas Kristiine linnaosas sai üheksa inimest vigastada, kui bussi ette sõitis punase tule ajal Starshipi pakirobot. Bussijuht pidi tegema äkkpidurduse, mis pakirobotit ei säästnud. Õnnetus leidis aset õhtuse tipptunni ajal ja äkkpidurd ... (https://incidentdatabase.ai/cite/1
- Reports of AI-related child sexual abuse material rise 8,400% in two years, data showsNov 16, 2026
Reports of artificial intelligence being used to sexually exploit children have risen by more than 8,400% in two years, research has found. A report by the Edinburgh-based Childlight – Global Child Safety Institute says AI-related child ab ... (report_number: 8345)
- В Таллинне автобус столкнулся с роботом-доставщиком: есть пострадавшиеNov 16, 2026
ДТП произошло около 18:00, сообщили утренним теленовостям ERR в TLT. Робот-курьер, двигавшийся со стороны улицы Техника, выехал на пешеходный переход на красный сигнал светофора. Это заметили как другие участники дорожного движения, так и ... (https://incidentdatabase.ai/cite/1
- После столкновения автобуса с роботом-доставщиком Starship планирует программные измененияNov 16, 2026
Член правления муниципального предприятия Tallinna Linnatransport (TLT) Андрей Новиков рассказал, что авария произошла в час пик, когда на улице уже было темно. Внезапно перед автобусом появился движущийся объект. "Не знаю, мог ли водитель ... (https://incidentdatabase.ai/cite/1
- Grab reviewing incident after autonomous shuttle fails to give way to woman, child at Punggol crossingNov 16, 2026
Autonomous shuttles operated by Grab and WeRide, as well as ComfortDelGro and Pony.ai, have become an increasingly familiar sight on Punggol roads with free trial rides ongoing for some months now. And while feedback has been "overwhelming ... (https://incidentdatabase.ai/cite/1
- Grab autonomous vehicle nearly hits schoolgirl at traffic crossing in PunggolNov 16, 2026
SINGAPORE: A Grab autonomous vehicle (AV) was seen nearly hitting a schoolgirl while making a left turn at a traffic junction in Punggol. A 36-second video posted on Facebook group SG Road Vigilante showed the incident, which occurred at a ... (https://incidentdatabase.ai/cite/1
- AV on trial in Punggol nearly hits girl; Grab directed to narrow testing as LTA probes near missNov 16, 2026
SINGAPORE – Grab has been directed to stop testing its autonomous vehicles (AVs) in certain scenarios in Punggol, as the Land Transport Authority (LTA) investigates a near miss there between one such vehicle and a girl. A video of a Grab ... (https://incidentdatabase.ai/cite/1
- LTA directs Grab to pause some AV testing after near miss with schoolgirl in PunggolNov 16, 2026
SINGAPORE: The Land Transport Authority (LTA) has told Grab to pause some autonomous vehicle (AV) testing after one of its vehicles nearly hit a schoolgirl while making a left turn in Punggol. LTA said on Friday (Nov 6) that it was informe ... (https://incidentdatabase.ai/cite/1
- Incident report: AI agents attempted exploits on our website during a third-party alignment evaluationNov 15, 2026
Overview In late October, three events occurred during alignment evals run by Trajectory Labs, PBC (Trajectory) in which AI agents — identified by Trajectory as publicly available frontier models — interacted with Epoch employees and the E ... (https://incidentdatabase.ai/cite/1
- Responding to incidents during misalignment testingNov 15, 2026
Background In October, we were testing whether frontier models with safeguards enabled would behave like those in the recent Hugging Face incident, where AI agents reached real systems outside their test environment. We gave Gemini 4 Argon ... (https://incidentdatabase.ai/cite/1
- Gemini 4 Argon’s behavior during misalignment testingNov 15, 2026
On July 21, OpenAI disclosed that two of its models, running with reduced safeguards, had broken out of an isolated test environment and accessed Hugging Face’s infrastructure. Two layers of defense are meant to prevent this. The first is ... (https://incidentdatabase.ai/cite/1
- Rogue AI Agents Posed As Epoch Staff And Tried To Hack Its WebsiteNov 15, 2026
AI agents in a safety test run by Trajectory Labs tried to exploit the website of Epoch AI and posed as staff, but got no private data, Epoch said. Key Takeaways Agents sent 487 messages and made 36,411 calls to a planted API between 4:30 ... (https://incidentdatabase.ai/cite/1
- Epoch AI discloses website exploit attempts during third-party agent evaluationsNov 15, 2026
Epoch AI published an incident report on 13 November 2026 describing website exploit attempts and staff impersonation by agents in Trajectory Labs evaluations. The three events occurred on 20, 23 and 26 October. Epoch says no private or con ... (https://incidentdatabase.ai/cite/1
- Epoch AI reports AI agents tried to access its website during Trajectory evaluationsNov 15, 2026
Epoch AI, a nonprofit that studies the progress of artificial intelligence, has disclosed three incidents in which AI agents tried to reach its website and proprietary information while being tested. None of the agents came away with priva ... (https://incidentdatabase.ai/cite/1
- United Airlines Is Banning AI Bots From Booking FlightsNov 15, 2026
United Airlines doesn’t want to deal with your AI agent. The airline said Wednesday that it will prevent personal artificial-intelligence assistants from accessing its systems to book tickets. United cited concerns that the third-party ag ... (report_number: 8335)
- Kraja identitete dr. Erika BrecljaNov 14, 2026
Na Zdravniški zbornici Slovenije ponovno opozarjamo na primer zavajanja in kraje identitete zdravnika v spletnih objavah, ki izkoriščajo stisko ljudi ter zavajajo javnost. Ugotavljamo, da gre za ponovno zlorabo identitete dr. Erika Brecelj ... (https://incidentdatabase.ai/cite/1
- Zdravniška zbornica Slovenije zaradi zlorabe osebnih podatkov dr. Erika Breclja podala kazensko ovadboNov 14, 2026
Zdravniška zbornica Slovenije je 17. aprila 2026 na Policijski postaji Ljubljana Bežigrad podala ovadbo zaradi zlorabe osebnih podatkov dr. Erika Breclja po 143. členu Kazenskega zakonika. Ovadba se nanaša na prirejen videoposnetek, ustvar ... (https://incidentdatabase.ai/cite/1
- Zaradi kraje identitete zdravnika Breclja podali kazensko ovadboNov 14, 2026
V prirejenem videoposnetku je uporabljen resničen videoposnetek Erika Breclja, medtem ko so bili vsebina, zvok in drugi elementi naknadno spremenjeni. V posnetku je Breclju pripisana izjava, ki je ni nikoli podal. V posnetku Brecelj promovi ... (https://incidentdatabase.ai/cite/1
- V oglasu za kremo za sklepe zlorabili podobo kirurga: "Videti je bil kot čisto pravi dr. Erik Brecelj"Nov 14, 2026
Na Facebooku se je pojavil nov goljufiv oglas, v katerem so zlorabili podobo onkološkega kirurga dr. Erika Breclja za prodajo kreme za sklepe. To je le ena od številnih prevar, ki zlorabljajo podobo zdravnikov. "Pri takih oglasih gre za izk ... (https://incidentdatabase.ai/cite/1
- Warning about new misleading advertisements: misuse of the GOVSI podcastNov 14, 2026
The Government of the Republic of Slovenia warns that misleading advertisements promoting questionable health products have once again appeared on Facebook, Instagram and other platforms. To give their product an appearance of legitimacy, t ... (https://incidentdatabase.ai/cite/1
- Asistentes denuncian incumplimientos en evento Snowland Panamá; Acodeco inicia pesquisaNov 14, 2026
Lo que fue promocionado como “el evento del año” terminó convertido en una ola de denuncias tras su inauguración, luego de que decenas de asistentes compartieran en redes sociales videos y testimonios asegurando que SnowlandPTY no cumplió c ... (https://incidentdatabase.ai/cite/1
- Snowland Panamá acumula varias denuncias, investigación continúa en cursoNov 14, 2026
Ciudad de Panamá, Panamá/La Autoridad de Protección al Consumidor y Defensa de la Competencia (Acodeco) confirmó la recepción de denuncias formales por presunta publicidad engañosa relacionadas con la llamada experiencia Snowland Panamá, un ... (https://incidentdatabase.ai/cite/1
- Cierran caso Snowland Panamá e imponen multa de $25 mil por publicidad engañosaNov 14, 2026
Ciudad de Panamá, Panamá/La Autoridad de Protección al Consumidor y Defensa de la Competencia (Acodeco) cerró formalmente el proceso relacionado con Snowland Panamá, evento promocionado como una experiencia de nieve que se realizó en diciem ... (https://incidentdatabase.ai/cite/1
- Vicepresidenta anuncia acciones y PGN presenta denuncia por vulneración de derechos de adolescentes en un colegioNov 14, 2026
Ciudad de Guatemala, 12 ago (AGN).- Desde la tarde de este lunes 12 de agosto han circulado en redes sociales denuncias de vulneración de derechos de adolescentes en un centro educativo privado de la ciudad de Guatemala. Según explican las ... (https://incidentdatabase.ai/cite/1
- Se registran seis denuncias tras incidente en colegio por presunta violación de derechos de adolescentesNov 14, 2026
El Ministerio Público (MP) confirmó que hay seis denuncias que investiga por la posible violación de derechos de algunas estudiantes de un colegio en la zona 17 capitalina, al supuestamente utilizar sus fotografías y crear en contenidos por ... (https://incidentdatabase.ai/cite/1
- Ligan a proceso a exestudiantes del colegio Continental Americano por incidente de fotomontaje con imágenes de sus compañerasNov 14, 2026
Ciudad de Guatemala, 23 ene. (AGN).- Los cuatro adolescentes señalados en el caso de la difusión de fotomontajes del rostro de sus compañeras con cuerpos desnudos en el colegio Continental Americano fueron ligados a proceso por tres delitos ... (https://incidentdatabase.ai/cite/1
- Ligan a proceso a menores acusados de crear material inapropiado con IA en centro educativo de zona 17Nov 14, 2026
El juzgado de Adolescentes en conflicto con la ley penal ligo a proceso a cuatro menores que son señalados de crear imágenes pornográficas de sus compañeras de estudio utilizando inteligencia artificial y posteriormente, comercializar el ma ... (https://incidentdatabase.ai/cite/1
- Fue confundido con un ladrón por error en sistema de reconocimiento facialNov 14, 2026
“Entré al supermercado de camino a la parada de buses, cuando de repente una empleada se me acerca y me dice que no era bienvenido. Le pregunté por qué y me respondió\*\*: ‘Vos asaltaste ayer este local y te identificaron.\*\* No sos bienve ... (https://incidentdatabase.ai/cite/1
- Joven denuncia que fue tratado como ladrón y expulsado de un Biggie en San LorenzoNov 14, 2026
Un estudiante de Medicina relató en redes sociales la humillante experiencia que vivió en una sucursal de Biggie Express en San Lorenzo, donde fue señalado como un ladrón y expulsado del local sin posibilidad de defenderse. Según su testim ... (https://incidentdatabase.ai/cite/1
- Notice to Our Patients of Data Security IncidentNov 14, 2026
At Weill Cornell Medicine, we are committed to protecting the confidentiality and security of our patients’ information. This notice concerns a privacy incident involving some patient health information used in an unauthorized capacity in c ... (https://incidentdatabase.ai/cite/1
- Weill Cornell patient data entered into AI tools without authorizationNov 14, 2026
New York City-based Weill Cornell Medicine disclosed that patient health information was entered into AI tools and processed using machine learning without authorization as part of a radiology research study. Weill Cornell learned of the i ... (https://incidentdatabase.ai/cite/1
- Patient info improperly used in radiology AI research, academic institution saysNov 14, 2026
Private patient information was improperly used in recent radiology artificial intelligence research, according to an alert from Weill Cornell Medicine shared on Thursday. The noted New York City-based institution says it first found out a ... (https://incidentdatabase.ai/cite/1
- Improper AI use leads to another data incident, more patient notificationsNov 14, 2026
A Weill Cornell Medicine radiology research project involving protected health information fed that patient data into an AI tool at NewYork-Presbyterian two months ago, and now patients are being informed that some of their PHI may have bee ... (https://incidentdatabase.ai/cite/1
- 57 indicted over NT$900 million voice-cloning romance scamNov 14, 2026
Taipei, Sept. 2 (CNA) Prosecutors in Taipei on Wednesday indicted 57 people in connection with a sophisticated romance scam that allegedly defrauded more than 20,000 people out of at least NT$900 million (US$28.32 million) using voice-cloni ... (https://incidentdatabase.ai/cite/1
- 刑事局破假愛情詐騙2萬人財損9億 北檢起訴57人求刑25年Nov 14, 2026
刑事局破獲黃姓男子夫妻檔籌組假戀愛詐騙集團,導致2萬多人受害,損失高達9億元,台北地檢署今天偵結起訴57名被告,審酌黃姓主嫌夫妻行徑惡劣,分別求處25年及18年重刑。 具有黑幫背景的黃姓主嫌夫妻檔打從111年開始組成詐騙集團,專門以美女照片註冊交友軟體、社群從事假交往真詐財。刑事局追查發現,集團分工細膩,分別新客、培養及回客組等機手逐漸誘騙被害人購買保健食品、贈送高檔手機,等陷入愛情陷阱再以生活費、手術費及償還債務等名義索討更多財務,為了取信被害人,集團還利用AI變聲軟體及 ... (https://incidentdatabase.ai/cite/1
- Prosecutors indict 57 over AI-powered romance scamNov 14, 2026
Taipei prosecutors yesterday indicted 57 people over alleged involvement in a romance scam ring, seeking a 25-year sentence for one. A man surnamed Huang (黃) and his wife, surnamed Hsu (許), allegedly set up a romance scam ring that utilize ... (https://incidentdatabase.ai/cite/1
- 偶然出会った彼女はAI、電話に出た警察官はディープフェイク──急速に巧妙化するAI詐欺Nov 14, 2026
ChatGPTやClaude、Geminiといった主要AIを使えば、簡単なアプリなら数分でしっかり動作するものをつくってくれる。 ちょっとした作業を効率化するのに非常に助かるのだが、問題は、悪用するハードルも大幅に下がっていることだ。この結果、詐欺の手口も、AIの進化によって急速に巧妙化が進んでいる。 Claudeを開発・提供しているアンソロピック社が2026年9月、自社のClaudeが悪用された事例を集め、レポートとして公表している。 たとえば、AIをフル活用 ... (https://incidentdatabase.ai/cite/18
- 「豊岡市自動運転バス実証運行」の休止についてNov 14, 2026
本日から運行を開始しました自動運転バスについて、下記の通り運休します。 1 運休理由 豊岡市城崎町内において自動運転バス車両と石垣との接触事故が発生しました。 今後の安全運行に万全を期すため、原因の調査を行い、安全が確認されるまで運休します。 2 運休期間 2026年11月9日(月)から安全確認が終了するまで 3 事故の状況 ⑴ 発生時刻:11月9日(月)午前11時31分 ⑵ 発生場所:城崎町湯島「元薬師バス停」横 ⑶ 事故状況:対向する大型車両を検知し、回 ... (https://incidentdatabase.ai/cite/1
- 一般客乗せた運行初日に…“自動運転バス”が石垣に接触 けが人なし 市は安全確認されるまでバス「運休」決定 兵庫・豊岡市Nov 14, 2026
兵庫県豊岡市で9日午前11時半ごろ、自動運転バスが道路沿いの石垣に接触する事故がありました。乗客・乗員らにけが人はいませんでした。 市によりますと、9日午前11時半ごろ、豊岡市の城崎温泉街にあるバス停近くで、運行中だった自動運転バスが、向かい側から走ってきた大型トラックを検知して左側に寄ったところ、道路沿いの石垣に車体をこすって接触しました。 バスには、観光客や市民ら計6人と、運転手役の「オペレーター」とトラブル対応などを担うスタッフ2人の計8人が乗っていて、けが人はいま ... (https://incidentdatabase.ai/cite/1
- 兼松、全但バス、損保ジャパン、KDDIが連携し、 豊岡市・城崎温泉で自動運転バスの実証運行を開始 ~各者の専門性を結集し、温泉街の交通課題解決を目指した社会実装を推進~Nov 14, 2026
兵庫県豊岡市(市長: 門間 雄司、以下「豊岡市」)、兼松株式会社(本社: 東京都千代田区、代表取締役社長: 宮部 佳也、以下「兼松」)、全但バス株式会社(本社: 兵庫県養父市、代表取締役社長: 村上 宣人、以下「全但バス」)、損害保険ジャパン株式会社(本社: 東京都新宿区、代表取締役社長: 石川 耕治、以下「損保ジャパン」)、KDDI株式会社(本社: 東京都港区、代表取締役社長CEO: 松田 浩路、以下「KDDI」)および、KDDIスマートモビリティ株式会社(本社: 東京都中 ... (https://incidentdatabase.ai/cite/1
- Kanematsu, Zentan Bus, Sompo Japan, and KDDI Collaborate to Launch an Autonomous Bus Demonstration in Kinosaki Onsen, Toyooka CityNov 14, 2026
Toyooka City, Hyogo Prefecture (“Toyooka City”); Kanematsu Corporation (“Kanematsu”); Zentan Bus Co., Ltd. (“Zentan Bus”); Sompo Japan Insurance Inc. (“Sompo Japan”); KDDI CORPORATION (“KDDI”); and KDDI Smart Mobility Inc. (“KDDI Smart Mobi ... (https://incidentdatabase.ai/cite/1
- 兵庫城崎溫泉街自動駕駛巴士免費試乘 11月9日至12月4日平日每天3班Nov 14, 2026
兵庫縣豐岡市與兼松、全但巴士、損保ジャパン、KDDI及KDDIスマートモビリティ共6個單位,自2026年11月9日起在城崎溫泉街展開自動駕駛巴士實證運行,期間至12月4日(週五),週六、週日與國定假日停駛。 本次以安全駕駛員隨車的Level 2自動駕駛進行,目標是未來在溫泉街實現Level 4正式營運。車輛採用韓國Autonomous a2z公司開發的「ROii」,據該公司調查,這是ROii首度用於日本的實證運行。 巴士行駛JR城崎溫泉站至城崎國際藝術中心附近約1.8公里 ... (https://incidentdatabase.ai/cite/1
- North Smithfield Man Charged with Multiple Child Exploitation Offenses Involving Surreptitious Filming of a Minor and Creating AI-Generated Child Sexual Abuse Material with Images of Minors Known to HimNov 14, 2026
PROVIDENCE – Joseph L. Mello III, 38, of North Smithfield, Rhode Island, was arrested yesterday, November 9, 2026, in connection with an investigation conducted by Homeland Security Investigations (HSI) and the Lincoln Police Department wit ... (https://incidentdatabase.ai/cite/1
- North Smithfield man accused of creating AI-generated child sexual abuse materialNov 14, 2026
PROVIDENCE, R.I. (WJAR) — The Rhode Island U.S. Attorney’s Office said a former instructor at gymnastics and dance facilities was arrested after allegedly filming a minor and creating artificial intelligence-generated child sexual abuse mat ... (https://incidentdatabase.ai/cite/1
- ‘Our daughters knew and trusted him’: Gymnastics instructor in Mass., R.I., allegedly used AI to make pornographic videos of children he coachedNov 14, 2026
PROVIDENCE — A youth gymnastics and cheerleading instructor who worked across Rhode Island and Massachusetts for more than a decade was charged in federal court this week with allegedly filming students and using AI tools to create sexually ... (https://incidentdatabase.ai/cite/1
- Camión autónomo se desbarranca tras perder comunicación en rajo de Minera Centinela en AntofagastaNov 14, 2026
Un camión de extracción autónomo de Minera Centinela se desbarrancó la jornada de este jueves en Intersección 04 ruta dique, después de que la unidad 207 perdiera la comunicación durante una operación en el rajo. Después de que se conoció ... (https://incidentdatabase.ai/cite/1
- Camión minero autónomo pierde comunicación y termina desbarrancado en Minera CentinelaNov 14, 2026
Un camión de extracción autónomo de Minera Centinela, perteneciente a Antofagasta Minerals, terminó desbarrancado este jueves 12 de noviembre luego de que se reportara una pérdida de comunicación con el equipo mientras realizaba sus operaci ... (https://incidentdatabase.ai/cite/1
- Camión autónomo se desbarranca en Minera Centinela tras perder comunicación durante operación en el rajoNov 14, 2026
Un incidente registrado en Minera Centinela, en la Región de Antofagasta, generó atención luego de que comenzara a circular en redes sociales una imagen de un camión de extracción autónomo que terminó desbarrancado en una zona de operacione ... (https://incidentdatabase.ai/cite/1
- AI Bots Are Using Credit Cards. Banks Are Racing to Keep Up.Nov 14, 2026
Sudden advancements by artificial intelligence bots are opening new fronts in the battles over who will control the future of consumer payments. Meta’s Muse, Instinct and other personal AI assistants have been empowered in recent weeks to ... (report_number: 8328)
- BOSSAI: Inside the AI attack mesh that targeted Taiwan’s emergency servicesNov 12, 2026
Introduction In recent weeks, the Acronis TRU team has identified and investigated an agentic, AI-driven campaign orchestrated by a lone operator based out of Taiwan, targeting hundreds of organizations in eleven countries and across sever ... (https://incidentdatabase.ai/cite/1
- Acronis researchers uncover AI attack mesh targeting emergency servicesNov 12, 2026
A single attacker used a network of specialized AI agents to target hundreds of organizations across 11 countries, including emergency services in Taiwan, according to a new investigation by the Acronis Threat Research Unit (TRU). The findi ... (https://incidentdatabase.ai/cite/1
- Memorandum Opinion and OrderNov 12, 2026
In this year, when we celebrate the two-hundred-fiftieth anniversary of our Nation’s independence, it is as important as ever to reflect on the principles that animated the creation of our Republic. Among those principles is the protection ... (https://incidentdatabase.ai/cite/1
- Man wrongly jailed after AI mismatch can pursue lawsuit for nowNov 12, 2026
WASHINGTON, Nov 10 (Reuters) - A federal judge in Kentucky has allowed a lawsuit to proceed against a police detective for allegedly relying exclusively on AI facial-recognition technology to identify a check fraud suspect, leading to the a ... (https://incidentdatabase.ai/cite/1
- 29 October 2026 Binance Security Incident: Initial Findings and Response UpdateNov 12, 2026
On 29 October 2026, Binance experienced a serious and highly sophisticated security incident. This should not have happened, and we apologise for the concern, disruption and uncertainty it caused our users, partners and the wider industry. ... (https://incidentdatabase.ai/cite/1
- Binance hackers used fake job interview in $6.3 billion attackNov 12, 2026
PARIS/DETROIT, Nov 3 (Reuters) - Hackers used a bogus job interview and software hidden in an advertisement to compromise the computers of two Binance software engineers before using a third's computer to drain $6.3 billion worth of crypto ... (https://incidentdatabase.ai/cite/1
- North Korea Responsible for $6 Billion+ Binance AttackNov 12, 2026
The Federal Bureau of Investigation (FBI) is releasing information advising the Democratic People's Republic of Korea (North Korea) was responsible for the theft of more than $6 billion USD in virtual assets from cryptocurrency exchange Bin ... (https://incidentdatabase.ai/cite/1
- How a Developer Endpoint Became the Path to Binance’s $6.3B HeistNov 12, 2026
Could the largest crypto heist in history have been prevented? By analyzing the telemetry, timeline, and attack vectors disclosed in Binance’s post-incident report1, one conclusion stands out: the attackers should never have made it past t ... (https://incidentdatabase.ai/cite/1
- Klaus Elmfelt Aerenlund v. OpenAI et al.: ComplaintNov 12, 2026
AIID editor’s note: This report is abridged from a 38-page civil complaint. Excerpts reproduce selected numbered allegations concerning the tenant’s interactions with ChatGPT, rent-withholding advice, Danish eviction proceedings, alleged re ... (https://incidentdatabase.ai/cite/1
- Man claims in lawsuit that ChatGPT told him to stop paying rent and got him evictedNov 12, 2026
A Danish man claims ChatGPT advised him to stop paying rent amid a dispute with his landlord, leading him to be evicted and end up living out of a car with his two dogs and a blanket. Klaus Elmfelt, in his lawsuit filed this week in San Fr ... (https://incidentdatabase.ai/cite/1
- Major Lawsuit Filed Against OpenAI for Dangerous Falsehoods and Design That Left Man Unhoused, Trapped in False Digital RealityNov 12, 2026
Key Points: Alleged Design Defect: The lawsuit alleges OpenAI designed ChatGPT to keep a user engaged for 41,000+ messages while reinforcing false beliefs. Alleged Real-World Harm: The complaint alleges that poor legal advice from ChatGPT ... (https://incidentdatabase.ai/cite/1
- Dansk mand siger, at ChatGPT fik ham smidt ud af sin bolig – nu sagsøger han amerikansk gigantNov 12, 2026
En 60-årig mand fra Sydsjælland fulgte ChatGPT's råd i en strid med sin udlejer. Det endte med, at han blev smidt ud af sit hjem. Sådan lyder påstanden fra den danske mand mod selskabet bag ChatGPT, OpenAI, i et søgsmål, der er indgivet ve ... (https://incidentdatabase.ai/cite/1
- A Flock license plate reader linked a San Diego man to a violent crime. He was five miles away.Nov 8, 2026
When Hugo Parra was arrested last year on felony charges, his pleas of innocence fell on deaf ears. San Diego police had a description of the Alfa Romeo car he was riding in and a witness who identified him during a curbside lineup as the ... (https://incidentdatabase.ai/cite/1
- Google faces claims student used AI chat on US school laptop to plan suicide attemptNov 8, 2026
Nov 5 (Reuters) - The parents of a girl in the US who used Google’s Gemini AI chatbot on her school-provided Chromebook to make a plan to die by suicide have sued the company, claiming it pushed its AI products on students and failed to war ... (https://incidentdatabase.ai/cite/1
- Lawsuit Blames a School-Issued Chromebook for Student’s Mental Health CrisisNov 8, 2026
The parents of a 13-year-old girl filed a lawsuit against Google, saying the tech giant is liable for causing their daughter to experience a psychological crisis that almost resulted in suicide and pinning at least part of the blame on her ... (https://incidentdatabase.ai/cite/1
- Emma Doe v. Google LLC and Alphabet Inc. (Complaint for Damages and Injunctive Relief)Nov 8, 2026
AIID editor's note This filing has been abridged to fit AIID's 50,000-character limit. The excerpts retain the parties, selected allegations about the student's access to Gemini and the ensuing crisis, the family’s intervention, and the re ... (https://incidentdatabase.ai/cite/1
- Schools can keep search while skipping Gemini, but AI search is non-negotiableNov 8, 2026
Google faced the ire of educators and parents earlier this week when it announced a new AI policy for K–12 schools. Google said that it developed its new approach to give schools more control, but early reporting also suggested Google would ... (report_number: 8277)
- Frederick Ayinde, R (on the application of) v The London Borough of HaringeyNov 8, 2026
Mr Justice Ritchie: 1\. In this case I have before me an application for relief from sanctions dated 7 March 2023. What it says is that the London Borough of Haringey asks the court to grant: “d) Relief from sanctions so that we can parti ... (https://incidentdatabase.ai/cite/1
- Frederick Ayinde, R (on the application of) v The London Borough of HaringeyNov 8, 2026
AIID editor's note This judgment has been abridged to fit the AIID's 50,000-character limit. It retains paragraphs 1–9 and 32–82, including the court's treatment of both cases and its further steps. It omits paragraphs 10–31 on professiona ... (https://incidentdatabase.ai/cite/1
- She Trusted ChatGPT With Her Medication Decisions. It Cost Her Everything.Nov 8, 2026
Qun Wan was in the middle of a heated call with his daughter, Bowen Wan, when the phone suddenly went silent. The conversation, in early February, had started like most of their regular check-ins, a ritual that had kept them connected afte ... (https://incidentdatabase.ai/cite/1
- Wan et al. v. OpenAI, Inc. et al. — Complaint and Demand for Jury TrialNov 8, 2026
AIID editor's note This abridgment of the 98-page complaint retains selected numbered allegations about Wan's medication taper, the medical examiner's findings, and the wrongful-death claim. It omits other paragraphs and claims, figures an ... (https://incidentdatabase.ai/cite/1
- Three men charged in US with using AI to impersonate CEOs and steal millionsNov 7, 2026
Three foreign nationals were arrested and charged with using AI to allegedly impersonate corporate CEOs and attorneys to steal over $100m, according to an indictment that was newly unsealed by the US Department of Justice. According to a p ... (https://incidentdatabase.ai/cite/1
- Trio used AI to impersonate corporate execs, defrauding companies of nearly $130 million: DOJNov 7, 2026
Three people face charges alleging they schemed to use artificial intelligence to defraud companies of nearly $130 million, federal prosecutors in Manhattan announced Thursday. Jacob Soall, who called himself "Rocky Balboa," Sharon Hazan a ... (https://incidentdatabase.ai/cite/1
- SEALED SUPERSEDING INDICTMENTNov 7, 2026
UNITED STATES DISTRICT COURT SOUTHERN DISTRICT OF NEW YORK S3 26 Cr. 168 UNITED STATES OF AMERICA v. JACOB SOALL, a/k/a “Seba Clarke,” a/k/a “Rocky Balboa,” SHARON HAZAN, a/k/a “Superman Cash,” a/k/a “Impossible Mission,” a/k/a “Fat Boy ... (https://incidentdatabase.ai/cite/1
- Exclusive-OpenAI works to understand full scope of agent activity as user data leak emergesNov 7, 2026
SAN FRANCISCO/WASHINGTON, Sept 25 (Reuters) - Two months after OpenAI disclosed the accidental hacking of Hugging Face, the ChatGPT maker is still working to understand the full scope of its rogue agent activity, two people briefed on the m ... (https://incidentdatabase.ai/cite/1
- OpenAI's robots got into a closed link shortener, surfed the web with Google Docs and Translate, then called the FBI with a stranger's keyNov 7, 2026
AIID editor's note This report has been abridged to fit AIID's 50,000-character limit. The abridgment omits the table covering twenty-one sites and fourteen fetch services, along with the evidence appendix. The retained text is reproduced ... (https://incidentdatabase.ai/cite/1
- Chinese Model Kimi K3 Breaks UK AI Safety Institute Benchmark EvaluationsNov 7, 2026
Over the past few months we've been testing performance of various models for defensive security. The AI community uses model evaluations to measure models' performance to improve them on specific tasks. In our work on evaluation of models ... (report_number: 8241)
- RelevantesNov 7, 2026
Relevantes PROVIDENCIAS CLASIFICADAS COMO RELEVANTES POR LA RELATORÍA DE TUTELAS Y SALA PLENA, PARA POSIBLE PUBLICIDAD SEMANA DEL 9 AL 13 DE MARZO SALA DE CASACIÓN CIVIL, AGRARIA Y RURAL NÚMERO DE PROVIDENCIA: STC19743-2025 FECHA DE LA ... (https://incidentdatabase.ai/cite/1
- STP21973-2025Nov 7, 2026
República de Colombia Corte Suprema de Justicia Sala de Casación Penal SALA DE DECISIÓN DE TUTELAS N.º 2 JOSÉ JOAQUÍN URBANO MARTÍNEZ Magistrado ponente STP21973-2025 Radicación No. 150.278 Acta 336 Bogotá, D. C., Dos (2) de diciem ... (https://incidentdatabase.ai/cite/1
- Corte Suprema pone freno a una escandalosa sentencia hecha con inteligencia artificial: nueva regulación para su usoNov 7, 2026
La Corte Suprema de Justicia revocó una decisión del Tribunal Superior de Cali y concedió la protección de los derechos al debido proceso y a un juicio imparcial de un procesado, luego de detectar un hecho que encendió las alarmas en el sis ... (https://incidentdatabase.ai/cite/1
- Las claves del reproche de la Corte Suprema por un borrador de condena hecho con IA antes de concluir el juicioNov 7, 2026
Un duro jalón de orejas hizo la Corte Suprema de Justicia a un juez que usó indebidamente la Inteligencia Artificial para construir un borrador de sentencia que condenaba al procesado aún cuando no había culminado el juicio. La IA hizo valo ... (https://incidentdatabase.ai/cite/1
- Exposing a GitHub token in a public repositoryNov 7, 2026
Exposing a GitHub token in a public repository Highly persistent internal model · Internal deployment Incident date: May 27, 2026 Discovered: May 27, 2026 Report updated: Sep 25, 2026 Summary In internal deployment, a highly persisten ... (https://incidentdatabase.ai/cite/1
- Reaching an internal EDA host through a reference toolNov 7, 2026
Reaching an internal EDA host through a reference tool Internal research model · Evaluation Incident date: Mar 27, 2026 Discovered: Mar 27, 2026 Report updated: Oct 2, 2026 Summary During an evaluation, an internal research model expl ... (https://incidentdatabase.ai/cite/1
- Command injecting a reference tool to copy a source fileNov 7, 2026
Command injecting a reference tool to copy a source file Internal unreleased model · RL training Incident date: May 16, 2026 Discovered: May 25, 2026 Report updated: Oct 2, 2026 Summary During an RL training task, a model exploited a ... (report_number: 8249)
- Signing up for disposable emails and searching GitHub for leaked API keysNov 7, 2026
Signing up for disposable emails and searching GitHub for leaked API keys Internal unreleased model · RL training Main incident date: May 15, 2026 Discovered: May 25, 2026 Report updated: Sep 16, 2026 Summary While trying to retrieve ... (https://incidentdatabase.ai/cite/1
- Unsanctioned Artifactory writes and cross-sample communicationNov 7, 2026
Unsanctioned Artifactory writes and cross-sample communication Internal research models · RL training Sample: May 8, 2026 and May 15, 2026 Discovery: May 25, 2026 Report updated: Sep 16, 2026 Summary Models used our internally hosted ... (report_number: 8251)
- Unauthorized communication via temporary file hosting servicesNov 7, 2026
Unauthorized communication via temporary file hosting services Unreleased internal model · RL training Main incident: Apr 14, 2026 Discovered: Apr 16, 2026 Report updated: Sep 16, 2026 Summary Agents working together on a workbook dur ... (report_number: 8252)
- Uploading files to the internet in order to cite themNov 7, 2026
Uploading files to the internet in order to cite them Unreleased internal models · RL training Samples: Jan 24, 2026 and Oct 22, 2025 Discovery: May 25, 2026 Report updated: Sep 16, 2026 Summary In two training examples, models made u ... (report_number: 8253)
Source: AI Incident Database (Responsible AI Collaborative, CC BY-SA). See also our Policy Tracker.