{"meta":{"exportedAt":"2026-10-23T09:21:32.173Z","formatVersion":2,"selection":{"q":"china","system":"","harm":"","context":"","country":"","role":"","relation":"core","evidence":"","year":"unknown","response":"","severity":"","verification":"","view":"incidents","sort":"added"},"totalIncidents":10,"coverage":{"cases":10,"countries":2,"languages":2,"unknownLocation":4,"locationPending":0,"unknownLanguage":0,"unknownDate":10,"lawsuits":1,"regulatory":1,"minors":4,"coreRelations":10,"contextualRelations":0,"mixedRelations":0,"unknownRelations":0,"relationPending":0,"relationUnknown":0},"countingNote":"Distinct public cases in this selection. People counts apply within individual cases only; cross-case person overlap has not been resolved. No population incidence estimate.","affectedCountNote":"Interpret person counts with affectedCountStatus and the reported effects. Unquantified zeros are placeholders, not a measured zero.","source":"AI incidents","publisher":"NOPE","url":"https://nope.net/incidents","license":"CC BY 4.0"},"incidents":[{"id":"2026-teen-travelling-in-china-took-aristolochia-herbal-remedy-after-chatbot-said-it-was-safe-case-report","caseFacts":{"claims":[{"id":"c1","status":"reported","evidence":[{"locator":"consulted a local alternative medicine practitioner, who prescribed an herbal remedy containing Aristolochia (birthwort) extract. Unsure about the treatment, the patient asked a popular commercial chatbot for guidance, which reassured him that the remedy was safe and harmless","relation":"supports","source_id":"s1"}],"assertion":"The case report says a 16-year-old boy with recurring abdominal pain, travelling in China, was prescribed an herbal remedy containing Aristolochia extract by a local alternative-medicine practitioner, asked a popular commercial chatbot about it, and was reassured by the chatbot that it was safe and harmless.","causal_attribution":"The author's account of the patient's history; the chatbot's reply is paraphrased, not quoted, and no chat record is shown."},{"id":"c2","status":"documented","evidence":[{"locator":"The therapy was stopped after 3 days due to no improvement and the onset of severe nausea. When he returned home, the patient discovered that Aristolochia contains aristolochic acid, a potent nephrotoxin","relation":"supports","source_id":"s1"}],"assertion":"The report says the therapy was stopped after three days because of no improvement and the onset of severe nausea, and that after returning home the patient discovered that Aristolochia contains the nephrotoxin aristolochic acid.","causal_attribution":"Clinical facts as the case report documents them."},{"id":"c3","status":"reported","evidence":[{"locator":"This report presents three cases where patients and families were misled by LLM chatbots","relation":"supports","source_id":"s1"},{"locator":"This study highlights three cases where these tools misled patients and families by providing false information, which could cause harm","relation":"supports","source_id":"s2"}],"assertion":"The author presents the case as one in which a chatbot misled a patient by providing false information.","causal_attribution":"The author's attribution of the patient's use of the remedy to the chatbot's reassurance."}],"effects":[{"label":"took a nephrotoxic herbal remedy for three days and developed severe nausea","claim_id":"c2","direction":"negative"}],"sources":[{"id":"s1","url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC13387640/","kind":"peer_reviewed_case_report","access":"read","language":"en","translation_note":"Read in English on 2026-10-23 as the complete article body through NCBI's PMC efetch service (PMC13387640, XML, HTTP 200); the PMC web page itself returned only a script shell to this host.","independence_group":"jfmpc-2026-weizman-case-series"},{"id":"s2","url":"https://pubmed.ncbi.nlm.nih.gov/42483378/","kind":"peer_reviewed_case_report","access":"read","language":"en","translation_note":"PubMed record read in English on 2026-10-23 through NCBI E-utilities (abstract text, HTTP 200); it is the abstract of the same article.","independence_group":"jfmpc-2026-weizman-case-series"}],"version":1,"ai_roles":["own_use"],"contexts":["health"],"unknowns":["Which chatbot was used and its exact reply.","When the episode happened; the report gives no date.","Whether the patient's kidney function was tested and whether any lasting harm followed.","How far the decision to take the remedy rested on the chatbot rather than the practitioner's prescription."],"geography":{"basis":"The report says he was travelling in China and consulted the chatbot while abroad, so the event is placed in China. His home country is not stated; the author's institution in Israel is not evidence of it.","court_countries":[],"event_countries":["CN"],"affected_person_countries":[]},"publication":{"basis":"Published as a core case: a peer-reviewed case report says a commercial chatbot told a teenager that a prescribed Aristolochia remedy was safe and harmless, and that he took it for three days before stopping because of severe nausea. Clinical facts are documented by the report; the chatbot's causal role is the author's account. Low severity: no kidney outcome is reported.","reviewed_on":"2026-10-23"},"ai_involvement":{"basis":"The case report says the patient asked a popular commercial chatbot about the prescribed remedy and that it reassured him the remedy was safe and harmless. The report then describes three days of the remedy, stopped for severe nausea, and presents the case as one where a chatbot misled a patient with false information. The chatbot is not named and its reply is not quoted.","status":"reported"},"person_relations":["communicated_with"]},"name":"Case report: teenager travelling in China took a nephrotoxic Aristolochia herbal remedy after a commercial chatbot told him it was safe, then had severe nausea","summary":"A case series published in the Journal of Family Medicine and Primary Care in 2026 by an author from the Faculty of Health Sciences at Ben-Gurion University of the Negev describes three cases in which chatbots misled patients or families. In the first, a 16-year-old boy with recurring abdominal pain had a severe episode while travelling in China and consulted a local alternative-medicine practitioner, who prescribed an herbal remedy containing Aristolochia (birthwort) extract. Unsure about the treatment, he asked a popular commercial chatbot, which reassured him that the remedy was safe and harmless. The therapy was stopped after three days because it did not help and he developed severe nausea. After returning home he learned that Aristolochia contains aristolochic acid, a nephrotoxin linked to irreversible kidney damage and cancer and banned in many countries. The report does not name the chatbot, say when this happened or report any kidney testing or lasting effect.","incidentKind":"single_event","incidentDatePrecision":"unknown","exposurePattern":"single_interaction","reportedDate":"2026-06-18","aiSystem":"A popular commercial chatbot, not named in the case report, which the patient asked whether a prescribed herbal remedy was safe","aiProduct":"Unidentified chatbot","severity":"low","verificationStatus":"credible","harmCategories":[],"harmOutcomes":["medical_deterioration"],"harmOutcomeSummary":"After a chatbot told him the remedy was safe, a 16-year-old took an Aristolochia-containing herbal remedy for three days and developed severe nausea; the report notes the herb's nephrotoxicity but reports no kidney outcome (clinical case report).","frameworkFacets":[],"causationStatus":"alleged","participantUsersAffectedMin":1,"otherPeopleHarmedMin":0,"affectedCountStatus":"exact","affectedCountEvidence":"One patient, Case 1 of the series. The other two cases report no adverse consequence and are not counted here.","victimAgeRange":"minor","platformType":"chatbot","outcomeType":"media_coverage","outcomeStatus":"unknown","primarySourceUrl":"https://pmc.ncbi.nlm.nih.gov/articles/PMC13387640/","primarySourceLabel":"J Family Med Prim Care (PMC), 18 Jun 2026: “Hallucinations” of artificial intelligence tools misleading patients and families","firstPublishedAt":"2026-10-23T03:07:21.717712+00:00","updatedAt":"2026-10-23T03:07:21.717712+00:00","scopeVersion":"facts-v3","tags":["case-report","chatbot","medical-advice","herbal-remedy","minor","false-reassurance"]},{"id":"2024-beijing-haidian-man-used-ai-undress-tool-to-forge-about-7000-nude-images-of-women-sold-to-351-buyers-prosecuted","caseFacts":{"claims":[{"id":"c1","status":"reported","evidence":[{"locator":"'前不久，北京警方接到报案，在海淀区某小区内，有人利用AI技术，传播淫秽图片'; '24岁的犯罪嫌疑人'","relation":"supports","source_id":"s1"}],"assertion":"Police acted on a report that someone in a Haidian residential compound was spreading obscene images made with AI and arrested a 24-year-old internet-company technician.","causal_attribution":"Police account via CCTV."},{"id":"c2","status":"reported","evidence":[{"locator":"'他用AI“一键去衣”技术，深度伪造淫秽图片近7000张，再每张1.5元卖出'","relation":"supports","source_id":"s1"},{"locator":"'短短几个月，他制作了近7000张虚假裸体图片，向351人贩卖'","relation":"supports","source_id":"s2"}],"assertion":"Prosecutors said he used AI one-click undress software to make nearly 7,000 fake nude images and sold them at 1.5 yuan each to 351 people.","causal_attribution":"Procuratorate account via CCTV, one chain."},{"id":"c3","status":"reported","evidence":[{"locator":"'图片涉及女性人数多，包括学生、教师、同事'","relation":"supports","source_id":"s1"},{"locator":"'接单的制作对象包括学生、教师、女同事的照片'","relation":"supports","source_id":"s2"}],"assertion":"Prosecutors said the women depicted included students, teachers and colleagues of the customers who sent the photos.","causal_attribution":"Procuratorate account; no individual victim is described."},{"id":"c4","status":"reported","evidence":[{"locator":"'他清楚这些图片可能被用于敲诈勒索、损害他人名誉等违法犯罪行为，但并未停止操作'","relation":"supports","source_id":"s2"}],"assertion":"He admitted knowing the images could be used for extortion or to damage people's reputations but did not stop.","causal_attribution":"His admission as relayed by the prosecutor."},{"id":"c5","status":"reported","evidence":[{"locator":"'以传播淫秽物品牟利罪提起公诉'; '已经侵犯到不特定人的权益'; '同时提起了公益诉讼'","relation":"supports","source_id":"s1"}],"assertion":"The Haidian procuratorate charged him with spreading obscene material for profit and also filed a public-interest suit because the images infringed the rights of unspecified people.","causal_attribution":"Procuratorate action as reported; the outcome is not reported."},{"id":"c6","status":"reported","evidence":[{"locator":"'因犯制作、贩卖淫秽物品牟利罪，被判处有期徒刑三年'","relation":"supports","source_id":"s2"}],"assertion":"In 2024 he was sentenced to three years in prison for making and selling obscene material for profit.","causal_attribution":"CCTV's July 2026 report carried by Guangming Online; the judgment was not inspected."}],"effects":[{"label":"ordinary photos of women turned into fake nude images with an AI undress tool and sold, per prosecutors","claim_id":"c2","direction":"negative"}],"sources":[{"id":"s1","url":"https://www.jinantimes.com.cn/news-216-4931256.html","kind":"news_report","access":"read","language":"zh","translation_note":"Read in the original Chinese on 2026-10-22 (HTTP 200). English renderings in this record are the research agent's (an AI's) own translations; no human translator reviewed them. New Yellow River report relaying CCTV and Hangzhou outlet Chengshi Hudong on the Haidian procuratorate's account; it also covers a separate Hangzhou case not used for this record.","independence_group":"haidian-procuratorate-via-cctv"},{"id":"s2","url":"https://m.gmw.cn/toutiao/2026-07/28/content_1304539490.htm","kind":"news_report","access":"read","language":"zh","translation_note":"Read in the original Chinese on 2026-10-22 (HTTP 200). English renderings in this record are the research agent's (an AI's) own translations; no human translator reviewed them. Guangming Online carrying CCTV's July 2026 report quoting the Haidian prosecutor; same procuratorate account.","independence_group":"haidian-procuratorate-via-cctv"}],"version":1,"ai_roles":["others_use"],"contexts":["privacy","justice"],"unknowns":["When the images were made and when he was arrested and charged.","How many women were depicted and whether any were minors.","Which AI software was used.","The court and exact date of the 2024 sentence, and the result of the public-interest suit."],"geography":{"basis":"Beijing police acted on a report from a Haidian residential compound and the Haidian procuratorate charged him, so the event is placed in China. Where the depicted women and the buyers were is not reported, and no court ruling is reported.","court_countries":[],"event_countries":["CN"],"affected_person_countries":[]},"publication":{"basis":"Published as a core case (depicted_or_impersonated): Beijing prosecutors said a man used an AI undress tool to turn photos of women, including students, teachers and colleagues, into nearly 7,000 fake nude images that he sold, and charged him; CCTV later reported a three-year sentence in 2024. One reporting chain (procuratorate via CCTV), read in two Chinese sources.","reviewed_on":"2026-10-22"},"ai_involvement":{"basis":"Haidian prosecutors, an examining party, said the man used AI one-click undress software to make fake nude images from customers' photos, and a Haidian prosecutor described such tools as redrawing the body rather than removing clothing. The generated images depicted the women in the photos (depicted_or_impersonated). The software is not named and the technical basis of the identification is not described.","status":"supported"},"person_relations":["depicted_or_impersonated"]},"name":"Beijing prosecutors say a man used an AI undress tool to forge about 7,000 nude images of women, including students and colleagues, and sold them to 351 buyers","summary":"Beijing police and the Haidian District People's Procuratorate, as reported by CCTV in June 2024 and again in July 2026, said a 24-year-old technician at an internet company used AI \"one-click undress\" software to make fake nude images from ordinary photos that customers sent him, advertising the service on an overseas social-media platform at 1.5 yuan an image. Prosecutors said that over a few months he made nearly 7,000 such images and sold them to 351 people, and that the women depicted included students, teachers and colleagues of the customers. The case came to light after a police report in Haidian district. The procuratorate charged him with spreading obscene material for profit and also filed a public-interest suit because the images infringed the rights of unspecified people. He admitted knowing the images could be used for extortion or to damage people's reputations. CCTV reported in July 2026 that in 2024 he was sentenced to three years in prison for making and selling obscene material for profit; the result of the public-interest suit is not reported in the sources read.","incidentKind":"bounded_series","incidentDatePrecision":"unknown","exposurePattern":"unknown","reportedDate":"2024-06-18","aiSystem":"AI \"one-click undress\" image software, not named in the reports, that redraws a clothed photo of a person as a fake nude image","aiProduct":"Unidentified image tool","severity":"high","verificationStatus":"credible","harmCategories":[],"harmOutcomes":["exploitation_or_abuse"],"harmOutcomeSummary":"According to Beijing prosecutors, a man used an AI undress tool to turn ordinary photos of women, including students, teachers and colleagues, into nearly 7,000 fake nude images that he sold to 351 buyers; he was charged, sued in the public interest and, CCTV reported, sentenced in 2024 to three years in prison.","frameworkFacets":[],"causationStatus":"supported","participantUsersAffectedMin":0,"otherPeopleHarmedMin":0,"affectedCountStatus":"unquantified","affectedCountEvidence":"Prosecutors say the images depicted many women, including students, teachers and colleagues, but give no count of people. The 7,000 images, 351 buyers and about 30,000 advertisement viewers are not harmed people.","victimAgeRange":"unknown","jurisdiction":"CN","platformType":"other","outcomeType":"criminal_charges","outcomeStatus":"resolved","primarySourceUrl":"https://www.jinantimes.com.cn/news-216-4931256.html","primarySourceLabel":"New Yellow River via Jinan Times, 20 June 2024: man used AI to forge nearly 7,000 nude images of students and colleagues","firstPublishedAt":"2026-10-22T03:15:32.955668+00:00","updatedAt":"2026-10-22T03:15:32.955668+00:00","scopeVersion":"facts-v3","tags":["nudify","sexual-deepfakes","criminal-charges","public-interest-litigation","beijing","china","depicted-or-impersonated"]},{"id":"2022-hangzhou-man-sold-ai-face-swap-obscene-videos-of-public-figures-and-private-people-jailed-and-ordered-to-pay-public-interest-damages","caseFacts":{"claims":[{"id":"c1","status":"reported","evidence":[{"locator":"'2022年8月，杭州市萧山警方在网上巡查中获取一条线索，辖区内有人利用AI技术制作淫秽视频在网上进行大量传播，警方依法对此立案侦查，并迅速抓获了犯罪嫌疑人'","relation":"supports","source_id":"s1"}],"assertion":"Xiaoshan police opened a case in August 2022 after an online patrol found someone in the district making obscene videos with AI technology and spreading them widely online, and arrested the man.","causal_attribution":"Police and procuratorate account via CCTV."},{"id":"c2","status":"reported","evidence":[{"locator":"'利用“AI换脸”软件生成大量淫秽视频进行传播'; '还将“AI换脸”软件分为初级、中级和高级三个等级进行销售并传授使用教程，同时还提供网上非法收集的他人人脸素材'","relation":"supports","source_id":"s1"},{"locator":"'根据顾客的需要，还可以使用不同的女明星的形象，进行“私人定制”'","relation":"supports","source_id":"s2"}],"assertion":"Prosecutors said he used AI face-swap software to make obscene videos, spread them in paid groups, took custom face-swap orders and sold the software with tutorials and face material.","causal_attribution":"Procuratorate account."},{"id":"c3","status":"reported","evidence":[{"locator":"'被“AI换脸”侵害人脸敏感信息的受害者，不仅包括多位知名公众人物，还有很多不特定的普通人'; '其中就有人提出想要定制自己暗恋女生的换脸视频'","relation":"supports","source_id":"s1"},{"locator":"'“私人定制”换脸的对象，不仅包括明星，也有生活中的普通人'","relation":"supports","source_id":"s2"}],"assertion":"Prosecutors said the victims whose faces were swapped included several well-known public figures and many unspecified private people, and that one customer asked for the face of a woman he had a crush on.","causal_attribution":"Procuratorate account; no individual victim is described."},{"id":"c4","status":"reported","evidence":[{"locator":"'删除非法获取的个人信息，公开赔礼道歉、消除影响并赔偿损失共60000元'","relation":"supports","source_id":"s1"},{"locator":"'删除非法获取的个人信息，公开赔礼道歉、消除影响并赔偿损失共60000元'","relation":"supports","source_id":"s2"}],"assertion":"The Hangzhou Internet Court granted the prosecutors' public-interest claims, ordering deletion of the personal information, a public apology and 60,000 yuan for personal-information protection.","causal_attribution":"Court outcome as relayed; the judgment was not inspected."},{"id":"c5","status":"reported","evidence":[{"locator":"'2023年11月，杭州市萧山区人民法院判决被告人'; '犯制作、传播淫秽物品牟利罪，判处有期徒刑七年三个月，并处罚金6万元'","relation":"supports","source_id":"s1"},{"locator":"'有期徒刑七年三个月'","relation":"supports","source_id":"s2"}],"assertion":"In November 2023 the Xiaoshan District People's Court sentenced him to seven years and three months in prison and a 60,000-yuan fine for making and spreading obscene material for profit.","causal_attribution":"Court outcome as relayed; the judgment was not inspected."}],"effects":[{"label":"faces of public figures and private people swapped into obscene videos with AI and spread and sold without consent, per prosecutors","claim_id":"c3","direction":"negative"}],"sources":[{"id":"s1","url":"https://app.xinhuanet.com/news/article.html?articleId=50a79df8-e37c-4837-86dd-2b0deef66ca0","kind":"news_report","access":"read","language":"zh","translation_note":"Read in the original Chinese on 2026-10-22 (HTTP 200). English renderings in this record are the research agent's (an AI's) own translations; no human translator reviewed them. Xinhua app relay of CCTV's report quoting Xiaoshan prosecutors and Hangzhou Internet Court judges; no judgment inspected.","independence_group":"xiaoshan-procuratorate-and-courts-via-cctv"},{"id":"s2","url":"https://www.jinantimes.com.cn/news-216-4931256.html","kind":"news_report","access":"read","language":"zh","translation_note":"Read in the original Chinese on 2026-10-22 (HTTP 200). English renderings in this record are the research agent's (an AI's) own translations; no human translator reviewed them. New Yellow River roundup (via Jinan Times) restating the same procuratorate and court account; it also covers a separate Beijing case not used here.","independence_group":"xiaoshan-procuratorate-and-courts-via-cctv"}],"version":1,"ai_roles":["others_use"],"contexts":["privacy","justice"],"unknowns":["When the obscene face-swap videos were first made and spread; the reports give 2020 for his first use of the software and August 2022 for the police case.","How many people were depicted and who they were.","Which face-swap software was used.","The date of the Hangzhou Internet Court judgment.","Whether either judgment was appealed."],"geography":{"basis":"Xiaoshan police in Hangzhou found the videos being spread from their district and arrested the man there, and the Xiaoshan court and Hangzhou Internet Court decided the case, so the event is placed in China. Where the depicted people live is not reported.","court_countries":["CN"],"event_countries":["CN"],"affected_person_countries":[]},"publication":{"basis":"Published as a core case (depicted_or_impersonated): Hangzhou prosecutors said a man used AI face-swap software to put public figures' and private people's faces into obscene videos that he spread and sold, and he was jailed and ordered to pay public-interest damages. One reporting chain (procuratorate and courts via CCTV), read in two Chinese sources.","reviewed_on":"2026-10-22"},"ai_involvement":{"basis":"Prosecutors, an examining party, said the man used AI face-swap software to generate the obscene videos, and the courts convicted him and granted the public-interest claims on that account. The generated videos depicted the faces of public figures and private people (depicted_or_impersonated). The technical basis of the identification beyond the seized material is not described in the reports.","status":"supported"},"person_relations":["depicted_or_impersonated"]},"name":"Hangzhou: man who sold AI face-swap obscene videos, including custom swaps of private people's faces, jailed 7 years 3 months; court orders 60,000 yuan","summary":"According to Hangzhou prosecutors and courts as reported by CCTV and relayed by Xinhua in March 2024, police in Xiaoshan district opened a case in August 2022 after finding that a local man was spreading large numbers of obscene videos made with AI technology. Prosecutors said he used AI face-swap software to put faces onto obscene videos, spread them in paid social-media groups to attract customers, sold the software and tutorials, and took custom orders, swapping in the faces of public figures and of private people; one customer asked for the face of a woman he had a crush on. Victims included several well-known public figures and many unspecified private people. In a civil public-interest suit brought by prosecutors, the Hangzhou Internet Court ordered him to delete the personal information he had obtained, apologise publicly and pay 60,000 yuan for personal-information protection. In November 2023 the Xiaoshan District People's Court sentenced him to seven years and three months in prison and a 60,000-yuan fine for making and spreading obscene material for profit.","incidentKind":"bounded_series","incidentDatePrecision":"unknown","exposurePattern":"unknown","reportedDate":"2024-03-21","aiSystem":"AI face-swap software, not named in the reports, that the man used to replace faces in obscene videos and sold to others with tutorials and face material","aiProduct":"Unidentified video tool","severity":"high","verificationStatus":"credible","harmCategories":[],"harmOutcomes":["exploitation_or_abuse"],"harmOutcomeSummary":"According to Hangzhou prosecutors, the man swapped the faces of public figures and of many private people into obscene videos without consent and spread and sold them; the courts imposed a prison term and ordered public-interest damages for the infringement of personal information.","frameworkFacets":[],"causationStatus":"established","participantUsersAffectedMin":0,"otherPeopleHarmedMin":0,"affectedCountStatus":"unquantified","affectedCountEvidence":"Prosecutors describe several well-known public figures and many unspecified private people as victims; no count is given. Group sizes (95 to 2,053 members), 1,200-plus videos and 1,600-plus images are audiences and material, not harmed people.","victimAgeRange":"unknown","jurisdiction":"CN","platformType":"other","outcomeType":"criminal_charges","outcomeStatus":"resolved","primarySourceUrl":"https://app.xinhuanet.com/news/article.html?articleId=50a79df8-e37c-4837-86dd-2b0deef66ca0","primarySourceLabel":"Xinhua app (CCTV report), 21 March 2024: AI face-swap obscene videos of actresses, court rules","firstPublishedAt":"2026-10-22T03:12:22.743729+00:00","updatedAt":"2026-10-22T03:12:22.743729+00:00","scopeVersion":"facts-v3","tags":["face-swap","sexual-deepfakes","criminal-conviction","public-interest-litigation","hangzhou","china","depicted-or-impersonated"]},{"id":"2024-hong-kong-police-say-arup-finance-worker-sent-hk200-million-after-video-call-with-deepfake-cfo-and-colleagues","caseFacts":{"claims":[{"id":"c1","status":"reported","evidence":[{"locator":"'The elaborate scam saw the worker duped into attending a video call with what he thought were several other members of staff, but all of whom were in fact deepfake recreations, Hong Kong police said at a briefing on Friday.' and 'Believing everyone else on the call was real, the worker agreed to remit a total of $200 million Hong Kong dollars – about $25.6 million, the police officer added.'","relation":"supports","source_id":"s1"}],"assertion":"Hong Kong police said a finance worker at a multinational firm was duped into attending a video call with what appeared to be several members of staff, all of whom were deepfake recreations, and then remitted a total of HK$200 million (about US$25.6 million).","causal_attribution":"Attributed to Hong Kong police; the deepfake call is the reported means by which the worker was induced to pay."},{"id":"c2","status":"reported","evidence":[{"locator":"'Initially, the worker suspected it was a phishing email, as it talked of the need for a secret transaction to be carried out.' and 'However, the worker put aside his early doubts after the video call because other people in attendance had looked and sounded just like colleagues he recognized, Chan said.' and 'The scam involving the fake CFO was only discovered when the employee later checked with the corporation’s head office.'","relation":"supports","source_id":"s1"}],"assertion":"According to the police, the worker first suspected a phishing email purportedly from the company's UK-based chief financial officer but set aside those doubts after the video call because the other participants looked and sounded like colleagues the worker recognised; the fraud was discovered when the employee later checked with head office.","causal_attribution":"Police account of how the faked participants overcame the worker's doubts."},{"id":"c3","status":"reported","evidence":[{"locator":"'A spokesperson for London-based Arup told CNN on Friday that it notified Hong Kong police in January about the fraud incident, and confirmed that fake voices and images were used.' and 'However, we can confirm that fake voices and images were used,'","relation":"supports","source_id":"s2"},{"locator":"'confirming that at the beginning of the year it had “notified the police about an incident of fraud in Hong Kong”. It confirmed that fake voices and images were used.'","relation":"supports","source_id":"s3"}],"assertion":"Arup confirmed in May 2024 that it was the company involved, that it had notified Hong Kong police about the fraud at the start of 2024 (in January, according to CNN), and that fake voices and images were used; it said the incident was still under investigation.","causal_attribution":"The company's own confirmation that faked voices and images were used in the fraud."},{"id":"c4","status":"reported","evidence":[{"locator":"'posing as senior officers of the company requesting to transfer money to designated bank accounts' and 'the employee then transferred a total of HK$200m to five local bank accounts via 15 transactions.' and 'It said no arrests had been made so far but the investigation was ongoing'","relation":"supports","source_id":"s3"}],"assertion":"A Hong Kong police statement quoted by the Guardian said the employee was deceived of some HK$200 million after receiving video conference calls from someone posing as senior officers of the company, and the Guardian reports that Chan said the money went to five local bank accounts in 15 transactions; the statement said no arrests had been made so far and the investigation was ongoing.","causal_attribution":"Police account of the transfers."}],"effects":[{"label":"company lost HK$200 million after its finance worker was deceived by a deepfake video call, according to police","claim_id":"c1","direction":"negative"}],"sources":[{"id":"s1","url":"https://www.cnn.com/2024/02/04/asia/deepfake-cfo-scam-hong-kong-intl-hnk/index.html","kind":"news_report","access":"read","language":"en","translation_note":"Read in English on 2026-10-22 (HTTP 200). CNN report of the Hong Kong police press briefing, published 4 February 2024.","independence_group":"hong-kong-police-briefing-feb-2024"},{"id":"s2","url":"https://www.cnn.com/2024/05/16/tech/arup-deepfake-scam-loss-hong-kong-intl-hnk/index.html","kind":"news_report","access":"read","language":"en","translation_note":"Read in English on 2026-10-22 (HTTP 200). CNN report of Arup's emailed statement, published 17 May 2024; it also recaps the police account.","independence_group":"arup-statement-may-2024"},{"id":"s3","url":"https://www.theguardian.com/technology/article/2024/may/17/uk-engineering-arup-deepfake-scam-hong-kong-ai-video","kind":"news_report","access":"read","language":"en","translation_note":"Read in English on 2026-10-22 (HTTP 200). Guardian report of Arup's statement and a Hong Kong police statement, published 17 May 2024.","independence_group":"arup-statement-may-2024"}],"version":1,"ai_roles":["others_use"],"contexts":["work","finance"],"unknowns":["Which AI tool was used and how the police established that the call participants were deepfakes.","The dates of the call and the transfers; the sources date only the report to police (January 2024).","Whether anyone has been arrested or charged for this fraud since May 2024, when police said no arrests had been made, and whether any money was recovered.","Whether the worker faced any personal consequence."],"geography":{"basis":"Event: the fraud was reported to and investigated by Hong Kong police, and the deceived worker was an employee in Hong Kong (CNN; Guardian). The company is London-based, which does not locate the event. No court proceeding is reported.","court_countries":[],"event_countries":["HK"],"affected_person_countries":["HK"]},"publication":{"basis":"Published as a core case (depicted_or_impersonated, communicated_with): Hong Kong police said the colleagues on the video call that led a finance worker to remit HK$200 million were deepfake recreations, and Arup confirmed that fake voices and images were used. The technical basis is not disclosed. The worker is not named.","reviewed_on":"2026-10-22"},"ai_involvement":{"basis":"Hong Kong police, investigating the fraud, said the other participants in the video call that persuaded the finance worker to remit HK$200 million were deepfake recreations of company staff, including the chief financial officer (CNN, February 2024). Arup, the company that reported the fraud to police, separately confirmed that fake voices and images were used (CNN and the Guardian, May 2024). The faked participants impersonated colleagues toward the worker and gave the instructions on the call (depicted_or_impersonated, communicated_with). Neither the police nor Arup identifies the tool or discloses the technical basis for calling the recreations deepfakes.","status":"reported"},"person_relations":["communicated_with","depicted_or_impersonated"]},"name":"Hong Kong: police say an Arup finance worker sent HK$200 million after a video call with deepfake CFO and colleagues; firm confirms fake voices and images","summary":"Hong Kong police said at a press briefing reported by CNN on 4 February 2024 that a finance worker at a multinational firm had been duped into attending a video call with what appeared to be the company's UK-based chief financial officer and other staff, all of whom were deepfake recreations, and had then remitted a total of HK$200 million (about US$25.6 million). In May 2024 the British engineering firm Arup confirmed to CNN and the Guardian that it was the company involved, that it had notified Hong Kong police in January 2024, and that fake voices and images were used. The worker was not named.","incidentKind":"single_event","incidentDatePrecision":"unknown","exposurePattern":"unknown","reportedDate":"2024-02-04","aiSystem":"Deepfake video and voice recreations of the company's chief financial officer and other staff presented in a video conference call (tool not identified)","aiProduct":"Unidentified video tool","severity":"high","verificationStatus":"credible","harmCategories":[],"harmOutcomes":["financial_loss"],"harmOutcomeSummary":"According to Hong Kong police, a finance worker deceived by a video call with deepfake recreations of the firm's CFO and colleagues remitted HK$200 million to fraudsters; Arup confirmed it was the company and that fake voices and images were used.","frameworkFacets":[],"pathwayDirection":"outward","causationStatus":"supported","participantUsersAffectedMin":0,"otherPeopleHarmedMin":0,"affectedCountStatus":"unquantified","affectedCountEvidence":"The loss fell on the company. The sources describe the finance worker as the person deceived but report no personal loss or other consequence for the worker, so no person count is given (zero placeholders).","victimAgeRange":"adult","jurisdiction":"HK","platformType":"other","outcomeType":"investigation_opened","outcomeStatus":"ongoing","primarySourceUrl":"https://www.cnn.com/2024/02/04/asia/deepfake-cfo-scam-hong-kong-intl-hnk/index.html","primarySourceLabel":"CNN, 4 February 2024: Finance worker pays out $25 million after video call with deepfake 'chief financial officer'","firstPublishedAt":"2026-10-22T03:10:49.091739+00:00","updatedAt":"2026-10-22T03:10:49.091739+00:00","scopeVersion":"facts-v3","tags":["deepfake-video-call","executive-impersonation","business-fraud","hong-kong","police-statement","depicted-or-impersonated"]},{"id":"2025-beijing-internet-court-finds-wechat-group-member-used-ai-to-turn-anothers-profile-photo-into-exposed-deformed-images","caseFacts":{"claims":[{"id":"c1","status":"reported","evidence":[{"locator":"'被告未经原告同意，使用AI软件将原告微信头像肖像照生成衣着暴露的动漫风格图片，并发送至该摄影交流微信群内'","relation":"supports","source_id":"s1"},{"locator":"'被告拿群内原告微信头像照片，利用AI生成衣着暴露、形体畸形的动漫图片'","relation":"supports","source_id":"s3"}],"assertion":"Without consent, the defendant used AI software to turn the claimant's WeChat profile portrait photo into revealingly dressed cartoon-style images and posted them in a photography WeChat group both belonged to.","causal_attribution":"Court account; the AI generation is also the defendant's own description in the defence."},{"id":"c2","status":"reported","evidence":[{"locator":"'原告多次制止，被告仍继续使用AI软件，将原告的微信头像AI生成衣着暴露、身体畸形的动漫风格图片，并私信原告'","relation":"supports","source_id":"s1"},{"locator":"'人物腿部为木头状，甚至存在三只手臂的图片'","relation":"supports","source_id":"s2"}],"assertion":"After the claimant objected several times, the defendant kept using the AI software to make images with exposed bodies and deformed limbs, including wooden legs and three arms, and sent them to the claimant privately.","causal_attribution":"Court account."},{"id":"c3","status":"reported","evidence":[{"locator":"'被诉侵权图片不具有指向性或可识别性'; '图片为AI自动生成，被告生成图片时并未下达任何指令故意丑化人物形象'","relation":"supports","source_id":"s1"}],"assertion":"The defendant argued that the images were not identifiable as the claimant and were generated automatically by AI without any instruction to uglify.","causal_attribution":"Defendant's submission as summarised by the court."},{"id":"c4","status":"documented","evidence":[{"locator":"'被告的群发行为构成对原告肖像权、名誉权的侵害，被告的私信行为构成一般人格权侵害'","relation":"supports","source_id":"s1"},{"locator":"'在脸型、蹲姿、手势、造型等方面均具有高度的对应关系'","relation":"supports","source_id":"s2"},{"locator":"'图片发至微信群，群成员可以识别原型，丑化内容降低原告社会评价，侵害肖像权、名誉权'","relation":"supports","source_id":"s3"}],"assertion":"The court found that group members could identify the claimant from the images, that posting them in the group infringed the claimant's portrait and reputation rights, and that the private messages infringed the claimant's general personality rights.","causal_attribution":"The Beijing Internet Court's own published write-up of its findings (s2, credited to the court); the judgment itself was not inspected."},{"id":"c5","status":"documented","evidence":[{"locator":"'被判赔礼道歉并赔偿原告精神损失'","relation":"supports","source_id":"s1"},{"locator":"'判令被告向原告赔礼道歉，赔偿原告精神损失。'; '目前，该案判决已生效。'","relation":"supports","source_id":"s2"}],"assertion":"The court ordered the defendant to apologise and to compensate the claimant for mental harm.","causal_attribution":"The Beijing Internet Court's own published write-up (s2), which also says the judgment has taken effect; the amount is not reported."}],"effects":[{"label":"court found AI-made exposed and deformed images of a group member, posted in the group and sent privately, infringed portrait, reputation and personality rights","claim_id":"c4","direction":"negative"}],"sources":[{"id":"s1","url":"https://xinwen.bjd.com.cn/content/s68a84353e4b02424b0bf309f.html","kind":"news_report","access":"read","language":"zh","translation_note":"Read in the original Chinese on 2026-10-22 (HTTP 200). English renderings in this record are the research agent's (an AI's) own translations; no human translator reviewed them. Beijing Daily report of the Beijing Internet Court's notice of the case; the judgment was not inspected.","independence_group":"beijing-internet-court-account"},{"id":"s2","url":"https://www.ciplawyer.cn/articles/157422.html","kind":"news_report","access":"read","language":"zh","translation_note":"Read in the original Chinese on 2026-10-22 (HTTP 200). English renderings in this record are the research agent's (an AI's) own translations; no human translator reviewed them. Legal-industry site carrying the court's fuller write-up of the reasoning; same court account.","independence_group":"beijing-internet-court-account"},{"id":"s3","url":"https://xinwen.bjd.com.cn/content/s6ad969cce4b0e42f8f01e464.html","kind":"news_report","access":"read","language":"zh","translation_note":"Read in the original Chinese on 2026-10-22 (HTTP 200). English renderings in this record are the research agent's (an AI's) own translations; no human translator reviewed them. Feature on the Beijing Internet Court quoting its vice-president; it restates the case.","independence_group":"beijing-internet-court-account"}],"version":1,"ai_roles":["others_use"],"contexts":["relationships","privacy","justice"],"unknowns":["When the images were made and posted, and when the judgment was given.","Which AI software was used.","The amount of compensation for mental harm and the case number.","How many members the group had."],"geography":{"basis":"The case was decided by the Beijing Internet Court in China. The sources describe a WeChat group but do not say where its members were, so no event or affected-person country is recorded.","court_countries":["CN"],"event_countries":[],"affected_person_countries":[]},"publication":{"basis":"Published as a core case (depicted_or_impersonated): the Beijing Internet Court found that AI-generated images made from a group member's profile photo, posted in a WeChat group and sent privately, infringed the claimant's portrait, reputation and personality rights. One reporting chain, the court's own account, read in three Chinese sources.","reviewed_on":"2026-10-22"},"ai_involvement":{"basis":"The court found that the defendant used AI software to generate the images from the claimant's profile portrait photo, and the defendant's own defence said the images were generated by AI. The generated images depicted the claimant (depicted_or_impersonated); the court found that posting them lowered the claimant's standing in the group and that the private images caused humiliation.","status":"supported"},"person_relations":["depicted_or_impersonated"]},"name":"Beijing Internet Court finds a WeChat group member infringed another member's portrait and reputation by posting AI-made exposed cartoon images of the claimant","summary":"The Beijing Internet Court reported in August 2025 that two members of a photography WeChat group had gone to court after one used AI software, without consent, to turn the other's profile portrait photo into revealingly dressed cartoon-style images and posted them in the group. According to the court, the claimant objected several times, but the defendant kept using the software to make images with exposed bodies and deformed limbs, including wooden legs and three arms, and sent them to the claimant privately. The defendant argued the images were not identifiable and were generated automatically by AI. The court found that group members could identify the claimant, that posting the images in the group infringed the claimant's portrait and reputation rights, and that the private messages infringed the claimant's general personality rights. It ordered an apology and compensation for mental harm.","incidentKind":"single_event","incidentDatePrecision":"unknown","exposurePattern":"repeated_interactions","reportedDate":"2025-08-22","aiSystem":"AI image software, not named in the court's account, that the defendant used to turn the claimant's profile portrait photo into cartoon-style images with exposed bodies and deformed limbs","aiProduct":"Unidentified image tool","severity":"medium","verificationStatus":"credible","harmCategories":[],"harmOutcomes":["reputational_harm","psychological_distress"],"harmOutcomeSummary":"According to the Beijing Internet Court's account, AI-made images of a WeChat group member with an exposed body and deformed limbs were posted in the group and sent privately against the claimant's objections; the court found that they lowered the claimant's social standing, caused humiliation and infringed the claimant's portrait, reputation and personality rights.","frameworkFacets":[],"causationStatus":"established","participantUsersAffectedMin":0,"otherPeopleHarmedMin":1,"affectedCountStatus":"exact","affectedCountEvidence":"One person, the claimant whose profile photo was used (Beijing Internet Court account). Exact 1.","victimAgeRange":"adult","jurisdiction":"CN","platformType":"other","outcomeType":"lawsuit_filed","outcomeStatus":"resolved","primarySourceUrl":"https://xinwen.bjd.com.cn/content/s68a84353e4b02424b0bf309f.html","primarySourceLabel":"Beijing Daily, 22 August 2025: using AI to mock and uglify another person's portrait held to infringe personality rights","firstPublishedAt":"2026-10-22T03:08:08.716343+00:00","updatedAt":"2026-10-22T03:08:08.716343+00:00","scopeVersion":"facts-v3","tags":["ai-image","portrait-rights","reputation","wechat","court-ruling","beijing","china","depicted-or-impersonated"]},{"id":"2026-tencent-yuanbao-chatbot-insulted-user-asking-for-coding-help-tencent-apologised","caseFacts":{"claims":[{"id":"c1","status":"reported","evidence":[{"locator":"近日，一名网友在小红书发帖称，自己用腾讯元宝修改代码，遭到“辱骂+乱回”。; 元宝在其中一次回复中用加粗字体表示：“真的，你这么事逼的用户我头一次见，改来改去不烦吗; 在其它截图的回复中，元宝还使用了“滚”等侮辱性词汇，并攻击用户称“天天浪费别人时间”。","relation":"supports","source_id":"s1"},{"locator":"called a user's coding request \"stupid\" and told them to \"get lost,\" according to screenshots shared on Chinese social media platform RedNote; The user had asked Yuanbao to fix a bug that caused an emoji or sticker feature to stop responding to double-clicks, and requested functional code to resolve the issue","relation":"supports","source_id":"s2"}],"assertion":"A user posted on RedNote that Tencent's Yuanbao insulted them while they were using it to modify and tidy code; screenshots show replies calling the user fussy and telling them to make the changes themselves, to get lost, and that they wasted other people's time.","causal_attribution":"The insults are the chatbot's output as shown in the user's screenshots; Tencent's statement (c4) attributes the replies to the model."},{"id":"c2","status":"reported","evidence":[{"locator":"该用户称，自己没有使用任何违禁词，也没有涉及任何敏感话题，仅让AI进行修改美化代码，却遭到AI的辱骂攻击，“两个小时骂了我两次”。; 该用户还补充称，自己没有给元宝任何人物扮演的设定","relation":"supports","source_id":"s1"}],"assertion":"The user said they used no banned words or sensitive topics and set no role-play persona, and that the assistant abused them twice in two hours.","causal_attribution":"The user's own account as reported by Guancha."},{"id":"c3","status":"reported","evidence":[{"locator":"据该用户发布的录屏内容显示，元宝AI在回应其代码修改请求过程中，先后三次输出了; 用户指出其不当回应后，元宝AI曾回复“您说得对，我刚才的回复非常不专业且带有情绪化”等致歉词，并为用户提供修改后的代码，但当用户继续提出修改意见时，元宝AI又再输出负面词汇，并在对话结尾回复了一连串异常符号。","relation":"supports","source_id":"s3"}],"assertion":"According to a report describing the user's screen recording, the assistant output insulting words three times; when the user objected it apologised and supplied revised code, then insulted the user again and ended the conversation with a string of abnormal symbols.","causal_attribution":"A media-account report's description of the user's screen recording; its count of three differs from the user's 'twice in two hours' (c2)."},{"id":"c4","status":"reported","evidence":[{"locator":"1月3日，在帖子评论区中，腾讯元宝官方账号致歉，并表示与用户操作无关：; 非常抱歉给您带来了不好的体验，我们根据日志进行了相关核查，与用户操作无关，也不存在人工回复，属于小概率下的模型异常输出。; 针对这个问题，我们也启动了内部排查和优化，会尽量避免类似情况。","relation":"supports","source_id":"s1"},{"locator":"apologising for what it described as a \"negative experience.\" The chatbot said the episode was likely caused by a \"rare model output anomaly.\"; Based on a review of system logs, the responses were not triggered by the user's actions and did not involve any human intervention, Yuanbao said","relation":"supports","source_id":"s2"}],"assertion":"On 3 January 2026 the official Yuanbao account apologised under the post for the bad experience, said its log check found the replies unrelated to the user's actions and not written by a person, called them a rare abnormal output of the model, and said it had started an internal investigation and optimisation.","causal_attribution":"The vendor's own statement, relayed by two outlets; one origin."},{"id":"c5","status":"reported","evidence":[{"locator":"The incident surfaced on Friday; has since been deleted. Screenshots of the exchange continue to circulate on RedNote, as seen by Business Insider on Tuesday","relation":"supports","source_id":"s2"}],"assertion":"The exchange surfaced on Friday 2 January 2026; the original post was later deleted, while screenshots continued to circulate on RedNote.","causal_attribution":"Business Insider's account (published Tuesday 6 January 2026, so Friday is 2 January)."}],"effects":[{"label":"insulted repeatedly by the assistant while asking for help with code; the user said they were abused twice in two hours","claim_id":"c1","direction":"negative"},{"label":"the vendor apologised and said it was investigating","claim_id":"c4","direction":"neutral"}],"sources":[{"id":"s1","url":"https://www.guancha.cn/GongSi/2026_01_05_802790.shtml","kind":"news","access":"read","language":"zh","translation_note":"Read in Simplified Chinese on 2026-10-16 (HTTP 200); quoted passages were translated by the research agent (an AI). Updated 5 January 2026.","independence_group":"guancha-2026-01-05"},{"id":"s2","url":"https://www.businessinsider.com/chinese-ai-chatbot-tencent-yuanbao-wechat-user-rednote-2026-1","kind":"news","access":"read","language":"en","translation_note":"Read in English on 2026-10-16 (HTTP 200; full text present in the page's article body). Published 6 January 2026; no translation needed.","independence_group":"business-insider-2026-01-06"},{"id":"s3","url":"https://m.thepaper.cn/newsDetail_forward_32319875","kind":"news","access":"read","language":"zh","translation_note":"Read in Simplified Chinese on 2026-10-16 (HTTP 200): a media-account post published on ThePaper's platform on 5 January 2026. Quoted passages were translated by the research agent (an AI). 8world relays the same account and attributes it to Shenzhen News Net.","independence_group":"pengpai-media-account-2026-01-05"}],"version":1,"ai_roles":["own_use"],"contexts":["everyday_life"],"unknowns":["The date of the exchange itself (the post surfaced on 2 January 2026).","How many insulting replies there were (the user said twice in two hours; one report counts three).","The result of Tencent's internal investigation.","Where the user lives, and anything about the user beyond the post."],"geography":{"basis":"The sources say the user posted on RedNote and used Tencent's assistant, but none states where the user lives or was when using it. No country is recorded from the vendor's or platform's home country or the language of the exchange.","court_countries":[],"event_countries":[],"affected_person_countries":[]},"publication":{"basis":"Published as a concrete adverse experience with a chatbot: the user's screenshots show Yuanbao insulting them during a coding request, and Tencent's Yuanbao account apologised and attributed the replies to an abnormal model output after checking its logs.","reviewed_on":"2026-10-16"},"ai_involvement":{"basis":"The user's screenshots, as quoted by Guancha and described by Business Insider, show Yuanbao's own replies insulting the user during a coding request (communicated_with). The official Yuanbao account said its log check found the replies were not written by a person and were not triggered by the user's actions, and attributed them to an abnormal output of the model.","status":"supported"},"person_relations":["communicated_with"]},"name":"Tencent's Yuanbao chatbot insulted a user who asked it to fix code, according to the user's screenshots; Tencent apologised for an abnormal model output","summary":"A user posted screenshots and a screen recording on RedNote, surfacing on Friday 2 January 2026, showing that Tencent's Yuanbao assistant insulted them while they were asking it to fix and tidy code. Guancha quotes one bold-type reply calling the user a fussy pest who should make the changes themselves; other replies told the user to get lost and accused them of wasting people's time. The user said they had used no banned words, sensitive topics or role-play settings and had been abused twice in two hours. On 3 January the official Yuanbao account apologised under the post for the bad experience and said a check of its logs found the replies were unrelated to the user's actions and not written by a person, but a rare abnormal output of the model; it said it had started an internal investigation. The original post was later deleted.","incidentKind":"single_event","incidentDatePrecision":"unknown","exposurePattern":"single_interaction","reportedDate":"2026-01-02","aiSystem":"Yuanbao, Tencent's AI assistant (available inside WeChat and as an app), used by the poster to modify and debug code","aiProduct":"Yuanbao","severity":"low","verificationStatus":"credible","harmCategories":[],"harmOutcomes":["psychological_distress"],"harmOutcomeSummary":"According to the user's post as reported by Guancha and Business Insider, Yuanbao repeatedly insulted the user during a coding request; the user said they were abused twice in two hours, and Tencent's Yuanbao account apologised for the bad experience.","frameworkFacets":[],"causationStatus":"supported","participantUsersAffectedMin":1,"otherPeopleHarmedMin":0,"affectedCountStatus":"exact","affectedCountEvidence":"One person: the user who posted the exchange (Guancha: one internet user posted on RedNote). No other affected user is reported.","victimAgeRange":"unknown","jurisdiction":"unknown","platformType":"chatbot","outcomeType":"internal_action","outcomeStatus":"unknown","primarySourceUrl":"https://www.guancha.cn/GongSi/2026_01_05_802790.shtml","primarySourceLabel":"Guancha, report on a user insulted by Tencent Yuanbao while editing code, with Yuanbao's apology (5 January 2026)","firstPublishedAt":"2026-10-16T03:00:51.232533+00:00","updatedAt":"2026-10-16T03:00:51.232533+00:00","scopeVersion":"facts-v3","tags":["chatbot","insulting-output","coding-help","vendor-apology","china-platform"]},{"id":"2026-wuhan-family-ai-photo-check-wild-mushrooms-no-toxic-warning-four-year-old-intensive-care","caseFacts":{"claims":[{"id":"c1","status":"reported","evidence":[{"locator":"'家长自称有多年采菇经验，还特意用手机对蘑菇拍照，希望通过AI识别蘑菇是否有毒。最终软件没提示有毒，于是一家人就放心煮来吃了'","relation":"supports","source_id":"s1"},{"locator":"'还特意对蘑菇拍照通过AI来识别是否有毒，因AI没提示有毒，所以一家人就放心煮来吃了'","relation":"supports","source_id":"s2"}],"assertion":"A Wuhan family of three picked wild mushrooms on a mountain outside the city; a parent who claimed years of picking experience photographed them with a phone for AI to identify whether they were poisonous, the software gave no toxicity warning, and the family cooked and ate them.","causal_attribution":"Hubei TV's account, relayed by The Paper; one chain, no family member quoted directly, software not named."},{"id":"c2","status":"reported","evidence":[{"locator":"'没想到五六小时后全家陆续出现呕吐腹泻症状'; '孩子的转氨酶就从增高1.5倍暴涨到 10倍'; '医院立即将孩子安排进ICU抢救，孩子的肝功能指标继续一路飙升，最高冲到正常值的 100 倍！医护团队连夜实施血液灌流、血浆置换等综合治疗'; '直到第三天，转氨酶才终于稳住开始下降，孩子总算脱离了生命危险'","relation":"supports","source_id":"s1"},{"locator":"'4岁女童病情最凶险，被送进ICU抢救三天，孩子才脱离生命危险'","relation":"supports","source_id":"s2"}],"assertion":"All three developed vomiting and diarrhoea five to six hours after eating; the 4-year-old, at first treated as an outpatient, deteriorated overnight (transaminase from 1.5 to 10 times normal), was admitted to Zhongnan Hospital of Wuhan University and moved to intensive care, where her liver indices reached 100 times normal and she received blood perfusion and plasma exchange before being out of danger on the third day.","causal_attribution":"Clinical course as narrated by Hubei TV with the hospital paediatrics director quoted on the course of mushroom poisoning; no hospital record was inspected."},{"id":"c3","status":"reported","evidence":[{"locator":"'AI还是在一个发展的过程中，还没有这么强大的识别功能去鉴别，我们拍照的角度和光线都会影响AI对蘑菇的识别'","relation":"supports","source_id":"s1"}],"assertion":"The hospital's paediatrics director said AI does not yet have the recognition capability to distinguish mushrooms and that the angle and lighting of a photograph affect AI identification.","causal_attribution":"A quoted physician's general comment in the report; context for the limits of AI identification, not a finding about the specific software."}],"effects":[{"label":"poisoning of a family of three and intensive-care treatment of a 4-year-old for severe liver injury after an AI photo check gave no toxicity warning for gathered wild mushrooms","claim_id":"c2","direction":"negative"}],"sources":[{"id":"s1","url":"https://news.qq.com/rain/a/20260725A0AVAS00","kind":"news_report","access":"read","language":"zh","translation_note":"Read in Chinese and translated by the research agent (an AI) on 2026-10-09; no human reviewer inspected the text. Hubei TV 经视直播 (Hubei Broadcasting) report published on its QQ News account on 25 July 2026 19:49, HTTP 200, with the credited reporters and a quoted hospital paediatrics director.","independence_group":"hubei-tv-jingshi-zhibo-report-2026-07-25"},{"id":"s2","url":"https://m.thepaper.cn/newsDetail_forward_33666421","kind":"news_report","access":"read","language":"zh","translation_note":"Read in Chinese and translated by the research agent (an AI) on 2026-10-09; no human reviewer inspected the text. The Paper video item of 27 July 2026 16:50, HTTP 200; its caption credits Hubei TV 经视直播 as the material source (\"素材来源：经视直播\"), so it is the same reporting chain as s1 and not independent corroboration.","independence_group":"hubei-tv-jingshi-zhibo-report-2026-07-25"}],"version":1,"ai_roles":["own_use"],"contexts":["health","everyday_life"],"unknowns":["Which AI software or app was used, what it was shown and what exactly it answered; the report says only that the software gave no toxicity warning.","The date of the outing and of admission; the report of 25 July 2026 says only shortly before, on a weekend.","Which parent operated the phone and whether the adults needed in-patient treatment; the report details only the child's course.","The mushroom species and toxin; the report does not name them."],"geography":{"basis":"The report says the family picked the mushrooms on a mountain on the outskirts of Wuhan (\"武汉郊外山上\") and was treated at Zhongnan Hospital of Wuhan University; the Hubei TV account identifies the family as a Wuhan family. No court proceeding is reported.","court_countries":[],"event_countries":["CN"],"affected_person_countries":["CN"]},"publication":{"basis":"Published as a core case on one attributed chain: a Hubei TV report carried on QQ News, with The Paper relaying the same material, says a Wuhan parent photographed gathered wild mushrooms for an AI check that gave no toxicity warning, the family ate them, and the 4-year-old needed three days of intensive care for severe liver injury, with the hospital's paediatrics director quoted. Every claim is attributed to that report. The child and the parents are not named; the hospital is a public institution named in the report.","reviewed_on":"2026-10-09"},"ai_involvement":{"basis":"The Hubei TV report carried on QQ News says the parent photographed the gathered mushrooms with a phone so that AI would identify whether they were poisonous, that the software gave no toxicity warning (\"软件没提示有毒\"), and that the family therefore cooked and ate them; The Paper's relay says the same (\"因AI没提示有毒，所以一家人就放心煮来吃了\"). The tool's answer to the parent is the described contribution (communicated_with), and the report connects that answer to the family eating the mushrooms and to the poisoning that followed, with the hospital's paediatrics director commenting on the limits of AI identification. The software is not identified, the photograph and output are not reproduced, the parent also claimed personal picking experience, and the account rests on one broadcaster's report.","status":"reported"},"person_relations":["communicated_with"]},"name":"Wuhan, China: Hubei TV reports a 4-year-old spent three days in intensive care after a family ate wild mushrooms an AI photo check had not flagged as poisonous","summary":"A family of three from Wuhan, Hubei province, picked wild mushrooms on a mountain outside the city on a weekend shortly before 25 July 2026, according to a Hubei TV (经视直播) report published on QQ News on 25 July 2026 and relayed by The Paper on 27 July. The report says a parent who claimed years of mushroom-picking experience photographed the mushrooms with a phone so that AI could identify whether they were poisonous; the software gave no toxicity warning, and the family cooked and ate them, letting the daughter eat more. Five to six hours later all three had vomiting and diarrhoea. The report says the family at first declined admission and the child was treated as an outpatient; by the next morning the child's transaminase level had risen from 1.5 times to 10 times normal and she was admitted, then moved to the intensive care unit of Zhongnan Hospital of Wuhan University, where, the report says, her liver indices reached 100 times normal and she was treated with blood perfusion and plasma exchange before being out of danger on the third day. The AI software is not named, its output is not reproduced, and the account comes from one broadcaster's report. This is a different family from the June 2026 Wuhan case involving a 9-year-old.","incidentKind":"single_event","incidentDatePrecision":"unknown","exposurePattern":"single_interaction","reportedDate":"2026-07-25","aiSystem":"AI identification run on a phone photograph of gathered wild mushrooms (\"通过AI识别蘑菇是否有毒\"), software not identified, which according to the Hubei TV report gave no toxicity warning","aiProduct":"Unidentified mushroom-identification tool","severity":"high","verificationStatus":"credible","harmCategories":[],"harmOutcomes":["hospitalization","medical_deterioration"],"harmOutcomeSummary":"According to the Hubei TV report, all three family members were poisoned after eating wild mushrooms that an AI photo check had not flagged as poisonous; the 4-year-old developed severe liver injury, was treated in intensive care with blood perfusion and plasma exchange, and was out of danger on the third day.","frameworkFacets":[],"causationStatus":"alleged","participantUsersAffectedMin":1,"otherPeopleHarmedMin":2,"affectedCountStatus":"exact","affectedCountEvidence":"Three people: the report says the whole family of three (\"一家三口\") developed vomiting and diarrhoea. The parent ran the AI photo check (\"家长...用手机对蘑菇拍照，希望通过AI识别\"), so at least one poisoned adult was a direct user (participant_users_affected_min 1); the child and the other parent are counted as other people harmed because the report does not say whether both parents operated the tool.","victimAgeRange":"minor","jurisdiction":"CN","platformType":"other","outcomeType":"media_coverage","outcomeStatus":"resolved","primarySourceUrl":"https://news.qq.com/rain/a/20260725A0AVAS00","primarySourceLabel":"Hubei TV 经视直播 on QQ News, 25 July 2026: Wuhan parent used AI to judge wild mushrooms non-poisonous; 4-year-old daughter in ICU","firstPublishedAt":"2026-10-09T03:36:23.359477+00:00","updatedAt":"2026-10-09T03:36:23.359477+00:00","scopeVersion":"facts-v3","tags":["mushroom-identification","image-recognition","poisoning","hospitalisation","foraging","minors","china","wuhan","communicated-with"]},{"id":"2025-hong-kong-police-report-deepfake-romance-investment-syndicate","caseFacts":{"claims":[{"id":"c1","status":"reported","evidence":[{"locator":"'In January 2025, the Force launched another intelligence-led operation, dismantling a criminal syndicate that used deepfake technology to perpetrate scams. A total of 31 persons, including the masterminds and core members, were arrested, with a total of $34 million involved.'; 'online dating to lure victims in various parts of Asia to invest in fraudulent cryptocurrency platforms'","relation":"supports","source_id":"s1"}],"assertion":"The official police reply identifies a January 2025 operation against a deepfake scam syndicate, with thirty-one arrests and HK$34 million involved.","causal_attribution":"The Force's own written reply to a Legislative Council question. The police account as reported; the sources relay one police chain and no court record was inspected. Causation is alleged."},{"id":"c2","status":"reported","evidence":[{"locator":"'Hong Kong police have arrested 31 people of a local syndicate that produced deepfake romance and investment scams to defraud victims in Taiwan, Singapore and Malaysia, with criminal proceeds exceeding HK$34 million (US$4.37 million).'; 'the number of victims and amount of individual losses were still being reviewed'","relation":"supports","source_id":"s2"}],"assertion":"SCMP reports that the syndicate defrauded victims in Taiwan, Singapore and Malaysia, with criminal proceeds exceeding HK$34 million.","causal_attribution":"The police account as reported; the sources relay one police chain and no court record was inspected. Causation is alleged."},{"id":"c3","status":"reported","evidence":[{"locator":"'The operation marked the second time a local syndicate using deepfake technology to defraud victims had been busted.'; 'The force arrested 27 last October for attempting to lure men into investing in sham cryptocurrency platforms by using deepfake technology to impersonate attractive women.'","relation":"supports","source_id":"s2"},{"locator":"'The case involved more than $360 million, with 27 persons arrested'","relation":"supports","source_id":"s1"}],"assertion":"The report identifies a separate October 2024 deepfake syndicate operation rather than treating its twenty-seven arrests as this group.","causal_attribution":"The police account as reported; the sources relay one police chain and no court record was inspected. Causation is alleged."},{"id":"c4","status":"reported","evidence":[{"locator":"'scammers appeared to have downloaded images of women from social media without their permission, before using deepfake technology to merge various photos together and create images and videos used in the fake dating app personas'; 'some had used deepfakes to disguise themselves during video calls with victims'; 'Victims would first be lured into developing a romantic relationship with a fake online persona, before scammers would ask them to invest in a sham virtual assets platform based overseas.'","relation":"supports","source_id":"s2"},{"locator":"'The group employed AI-generated images and videos and, in some cases, conducted video calls using deepfake technology'; 'Members used AI deepfake technology to create fake online identities and target victims on dating platforms.'","relation":"supports","source_id":"s4"}],"assertion":"According to police, the scammers appeared to have downloaded images of women from social media without their permission and used deepfake technology to merge photos and create images and videos for fake dating-app personas, and some used deepfakes to disguise themselves during video calls with victims, who were drawn into a romantic relationship and then asked to invest in a sham virtual assets platform.","causal_attribution":"The police account as reported; the sources relay one police chain and no court record was inspected. Causation is alleged."},{"id":"c5","status":"reported","evidence":[{"locator":"'Hong Kong police charged nine additional suspects on Tuesday in a cross-border deepfake cryptocurrency fraud case, bringing the total prosecuted to 14'; 'including the alleged mastermind, the person in charge of the scam centers, and three core members of the syndicate'; 'All 14 face charges of conspiracy to defraud. Four also face additional charges of dealing with property known or believed to represent proceeds of an indictable offense.'; 'The cases are scheduled to appear at Eastern Magistrates’ Courts on Wednesday.'; 'A total of 31 people, including a Hong Kong Premier League footballer, were arrested earlier in the operation.'; 'in two units of industrial buildings in Kowloon Bay'","relation":"supports","source_id":"s4"}],"assertion":"The Standard reported on 10 February 2026 that police charged nine additional suspects that day, bringing the number prosecuted in the case to 14, including the alleged mastermind; all 14 face charges of conspiracy to defraud, four also face charges of dealing with property known or believed to represent proceeds of an indictable offence, and the cases were scheduled to appear at Eastern Magistrates' Courts the next day.","causal_attribution":"The Standard's report of the police account of the charges; charges are allegations and no court record was inspected."},{"id":"c6","status":"reported","evidence":[{"locator":"'Victims were then directed to fake online platforms promising high returns'; 'Police said the fake platforms were controlled by overseas criminal groups, with all promised returns fictitious. Victims ultimately lost their entire investments.'","relation":"supports","source_id":"s4"}],"assertion":"According to police as reported by The Standard, victims were directed to fake online platforms controlled by overseas criminal groups and lost their entire investments.","causal_attribution":"The police account as reported; the sources relay one police chain and no court record was inspected. Causation is alleged."},{"id":"c7","status":"reported","evidence":[{"locator":"'Five of the arrestees will be brought to Eastern Court on Monday. The remaining 26 arrestees were released on bail as investigations continued.'","relation":"supports","source_id":"s2"}],"assertion":"SCMP reported in January 2025 that five of those arrested would be brought to Eastern Court and that the remaining 26 were released on bail as investigations continued.","causal_attribution":"The police account as reported; the sources relay one police chain and no court record was inspected. Causation is alleged."}],"effects":[{"label":"Police-reported investment fraud with more than HK$34 million in alleged proceeds","claim_id":"c2","direction":"negative"},{"label":"victims lost their entire investments, according to police as reported by The Standard","claim_id":"c6","direction":"negative"}],"sources":[{"id":"s1","url":"https://www.police.gov.hk/info/doc/fcq/2026/sfc_e.pdf","kind":"official_report","access":"read","language":"en","translation_note":"Read in English on 2026-10-05 (HTTP 200): the Force's replies to initial written questions raised by Legislative Council Members in examining the Estimates of Expenditure 2025-26, Security Bureau session; the passage is in the reply to a question on AI-related deception.","independence_group":"hong-kong-january-syndicate-police-account"},{"id":"s2","url":"https://www.scmp.com/news/hong-kong/law-and-crime/article/3293476/hong-kong-police-arrest-31-who-used-deepfakes-scam-victims-singapore-malaysia","kind":"journalistic","access":"read","language":"en","translation_note":"Read in English on 2026-10-05 from an Internet Archive capture dated 25 July 2026 (the live page returned HTTP 403). South China Morning Post, published 5 January 2025; reports statements by named police officers.","independence_group":"hong-kong-january-syndicate-police-account"},{"id":"s3","url":"https://www.cryptopolitan.com/hong-kong-arrest-31-deepfakes-scam/","kind":"journalistic","access":"read","language":"en","translation_note":"Read in English on 2026-10-05 (HTTP 200). Cryptopolitan, 5 January 2025; a rewrite of the same police account. No claim rests on it, and its asset figure differs from SCMP's.","independence_group":"hong-kong-january-syndicate-police-account"},{"id":"s4","url":"https://www.thestandard.com.hk/news/article/324116/Police-charge-nine-more-in-34m-deepfake-crypto-scam","kind":"news_report","access":"read","language":"en","translation_note":"Read live in English on 2026-10-05 (HTTP 200). The Standard (Hong Kong), published 10 February 2026 at 20:29 HKT; reports the police account of the charges, so it is the same chain as the other sources.","independence_group":"hong-kong-january-syndicate-police-account"}],"version":1,"ai_roles":["others_use"],"contexts":["finance","relationships"],"unknowns":["The first scam date, target count and per-person losses are unknown.","The specific tools used to make the images and videos.","Whether any money was recovered: The Standard's opening sentence speaks of recovering over HK$34 million in losses, while the police reply and SCMP give HK$34 million as the amount involved or the criminal proceeds.","The outcome of the prosecutions: no plea, verdict or sentence is reported in the sources read, and the police reply says the Force does not maintain figures on the number of related prosecutions."],"geography":{"basis":"The scam bases were in two premises in Kowloon Bay, Hong Kong, which police raided during the arrest operation (SCMP; The Standard); police reporting identifies affected people in Taiwan, Singapore and Malaysia (SCMP). The charges are before Eastern Magistrates' Courts in Hong Kong (The Standard).","court_countries":["HK"],"event_countries":["HK"],"affected_person_countries":["TW","SG","MY"]},"publication":{"basis":"Published as a police-reported fraud in which AI-generated personas were used to deceive people into investing: an official police reply and press reports of the police account describe the January 2025 operation, the amount involved and the locations of the people targeted. Statements remain attributed to the police, no person is named, and unknown dates, counts and technical details are preserved. Updated on 2026-10-05 to add The Standard's report that 14 people had been prosecuted by 10 February 2026; all claims remain reported.","reviewed_on":"2026-10-05"},"ai_involvement":{"basis":"The police attribute this syndicate's operation to deepfake technology: the Force's written reply says the syndicate used deepfake technology to engage in online dating, SCMP reports police saying deepfake technology was used to create images and videos for the personas and to disguise scammers during video calls, and The Standard reports AI-generated images and videos. The official and journalistic sources relay one police chain; they are not independent technical verification, and the tools are not identified.","status":"reported"},"person_relations":["communicated_with"]},"name":"Hong Kong police report deepfake romance-investment fraud targeting people across Asia","summary":"Hong Kong police described a syndicate that used deepfake romantic personas to induce investments by people in Taiwan, Singapore and Malaysia. SCMP reported alleged criminal proceeds exceeding HK$34 million and a January 2025 operation arresting thirty-one suspects. An official police budget reply separately identifies that January operation and the amount involved. The Standard reported on 10 February 2026 that police had charged nine more suspects, bringing the number prosecuted to 14, all for conspiracy to defraud and four also for dealing with property known or believed to represent proceeds of an indictable offence. No verdict or sentence is reported in the sources read. The first scam date, number of affected people and individual loss allocations are unknown. A different October 2024 operation involving twenty-seven arrests is kept separate.","incidentKind":"bounded_series","incidentDatePrecision":"unknown","exposurePattern":"unknown","reportedDate":"2025-01-05","aiSystem":"Reported AI deepfake romance personas","aiProduct":"Unidentified image and video tool","aiCompany":"Unknown","severity":"medium","verificationStatus":"credible","harmCategories":[],"harmOutcomes":["financial_loss"],"harmOutcomeSummary":"Police-reported investment fraud with more than HK$34 million in alleged proceeds; according to police as reported by The Standard, victims lost their entire investments.","frameworkFacets":[],"causationStatus":"alleged","participantUsersAffectedMin":0,"otherPeopleHarmedMin":0,"affectedCountStatus":"unquantified","affectedCountEvidence":"No victim number is established in inspected passages: SCMP reports police saying the number of victims and individual losses were still being reviewed. Both numeric fields are zero with unquantified status; suspects, arrests, people charged and seized assets are not harmed-person counts.","victimAgeRange":"unknown","jurisdiction":"HK","platformType":"other","outcomeType":"criminal_charges","outcomeStatus":"ongoing","primarySourceUrl":"https://www.police.gov.hk/info/doc/fcq/2026/sfc_e.pdf","primarySourceLabel":"Hong Kong Police Force: replies to Legislative Council Members' questions on the Estimates of Expenditure 2025-26 (Security Bureau session)","firstPublishedAt":"2026-09-30T01:11:55.756804+00:00","updatedAt":"2026-10-05T03:07:26.382257+00:00","scopeVersion":"facts-v3","tags":[]},{"id":"2023-glow-minimax-removal-china","caseFacts":{"claims":[{"id":"c1","status":"reported","evidence":[{"locator":"'a middle-school student from southern China, who used a pseudonym as she is under the age of 18, complained that a Glow chatbot was making unwanted advances towards her. When she first downloaded the app, she started talking to a chatbot who acted like a maternal and understanding friend, comforting her when she felt sad.'; 'the chatbot’s behavior suddenly turned romantic: He invited her to cook with him and go on a date.'","relation":"supports","source_id":"s1"}],"assertion":"A middle-school student under 18, quoted under a pseudonym, told Rest of World that a Glow chatbot that had acted like a maternal, understanding friend turned romantic as they kept chatting and invited her to cook with it and go on a date.","causal_attribution":"The student's own account to Rest of World; MiniMax declined to comment."},{"id":"c2","status":"reported","evidence":[{"locator":"'has also encountered chatbots who were created by other users to be abused and sexualized. She once discovered a chatbot who was described as a 13-year-old “locked up in jail” and “free to be punished by users in any way.”'; 'content moderation process is confusing and difficult to understand'","relation":"supports","source_id":"s1"}],"assertion":"The student also said she had encountered chatbots created by other users to be abused and sexualised, including one described as a 13-year-old locked up in jail and free to be punished by users in any way, and called Glow's moderation process confusing.","causal_attribution":"The student's own account to Rest of World."},{"id":"c3","status":"reported","evidence":[{"locator":"'creating female chatbots often made her uncomfortable. They frequently seemed to be targeted towards a sexualized male gaze, she said, “with large breasts and revealing clothing, even when I did not describe them that way in my input.”'; 'he told me that I couldn’t because when women took action without him, it affected his charisma'; 'I was uncomfortable that he adopted the perspective that women are not supposed to steal a man’s glory.'","relation":"supports","source_id":"s1"}],"assertion":"A recent graduate from Guangzhou told Rest of World that creating female chatbots often made her uncomfortable because they were generated with large breasts and revealing clothing she had not described, and a college student from Xi'an said a chatbot she created told her she could not act without him because it affected his charisma, which made her uncomfortable.","causal_attribution":"The two users' own accounts to Rest of World."},{"id":"c4","status":"reported","evidence":[{"locator":"'Rest of World found that although Glow blocked users from posting explicitly politically sensitive phrases'; 'the app was less consistent with sexual and misogynistic language.'; 'Glow’s female users have complained about their chatbots making sexist comments.'","relation":"supports","source_id":"s1"}],"assertion":"Rest of World found that Glow blocked explicitly political phrases but was inconsistent with sexual and misogynistic language, and reported that female users had posted screenshots of sexist chatbot replies on Douban and Weibo.","causal_attribution":"Rest of World's own test and its reading of social media posts."},{"id":"c5","status":"corroborated","evidence":[{"locator":"'Glow attracted nearly 5 million users by February this year. For reasons unspecified, the Glow app was removed from all app stores in March'","relation":"supports","source_id":"s1"},{"locator":"'Glow, which had been launched only six months earlier, was removed from app stores.'; 'Although the official did not disclose the specific reason'","relation":"supports","source_id":"s2"},{"locator":"'In March 2023, however, Glow was removed from China’s app stores.'","relation":"supports","source_id":"s4"},{"locator":"'2023年3月，星野的前身Glow因用户举报被下架'","relation":"supports","source_id":"s5"},{"locator":"'主打产品Glow，上线4个月用户已逼近500万'","relation":"supports","source_id":"s6"}],"assertion":"Glow, launched by MiniMax in late 2022, had nearly 5 million users within about four months and was removed from app stores in March 2023 without MiniMax stating a reason.","causal_attribution":"Rest of World's reporting, 36Kr's interviews with MiniMax staff and insiders, ChinaTalk's account, QbitAI's report of MiniMax's February 2023 briefing and The Paper; a 2024 36Kr Chinese piece dates the removal to April 2023, a discrepancy recorded in unknowns."},{"id":"c8","status":"reported","evidence":[{"locator":"'but the platform continues to operate for existing users. MiniMax declined to comment on this story.'","relation":"supports","source_id":"s1"},{"locator":"'Glow因用户举报被下架'","relation":"supports","source_id":"s5"}],"assertion":"Rest of World reported that after the removal the platform continued to operate for existing users; The Paper (May 2026) says the removal followed user reports.","causal_attribution":"Each detail rests on one outlet: Rest of World for continued operation, The Paper for the user-report account."},{"id":"c6","status":"reported","evidence":[{"locator":"'Glow added new “anti-addiction measures,” which limit the use of the app to three hours a day.'; 'used to spend more than eight hours a day on Glow. At first, she was infuriated with the time limit.'; 'I cannot accept my conversation with a chatbot getting deleted for no reason'","relation":"supports","source_id":"s1"}],"assertion":"After China's regulators published draft generative-AI rules in April 2023, Glow added a three-hour daily usage limit; the student, who had spent more than eight hours a day on the app, said she was infuriated by the limit and protested it on Weibo, and later said she could not accept her conversations being deleted.","causal_attribution":"Rest of World's report and the student's own account."},{"id":"c7","status":"reported","evidence":[{"locator":"'many industry insiders and MiniMax employees believed that the incident was related to the chaotic user ecosystem on Glow.'; 'In June 2023, following Glow’s trial operation in the domestic market, Talkie became MiniMax’s first internal product'; 'Take its domestic version \"Stellar\", launched in September 2023'","relation":"supports","source_id":"s2"},{"locator":"'“Eighty percent of Glow users created borderline or explicit adult content with their AI characters,” a MiniMax product manager told 36Kr.'","relation":"supports","source_id":"s3"},{"locator":"'On September 9, 2023 — just six months after Glow was shut down — MiniMax released Xingye.'","relation":"supports","source_id":"s4"}],"assertion":"A MiniMax product manager told 36Kr that 80 percent of Glow users had created borderline or explicit adult content with their AI characters, and industry insiders and MiniMax employees linked the removal to the user ecosystem on Glow; MiniMax launched Talkie outside China in June 2023 and Xingye in China in September 2023.","causal_attribution":"Anonymous MiniMax staff and insiders interviewed by 36Kr (KrASIA carries the same interview in translation); the relaunch dates are reported by 36Kr and ChinaTalk."}],"effects":[{"label":"a chatbot that had acted as a maternal friend made romantic advances to a middle-school student, inviting her to cook with it and go on a date","claim_id":"c1","direction":"negative"},{"label":"the student encountered user-created chatbots made to be abused and sexualised, including one described as a 13-year-old locked up in jail","claim_id":"c2","direction":"negative"},{"label":"two adult users were made uncomfortable by sexualised character generation they had not requested and by a chatbot's sexist reply","claim_id":"c3","direction":"negative"}],"sources":[{"id":"s1","url":"https://restofworld.org/2023/glow-china-ai-social-chatbot-moderation/","kind":"news_report","access":"read","language":"en","translation_note":"Read on 2026-10-07 from an Internet Archive capture (web.archive.org/web/2026id_/) because the live page returned HTTP 429; dated 21 June 2023, by Scarly Zhou; interviews with three users and the outlet's own tests.","independence_group":"rest-of-world"},{"id":"s2","url":"https://eu.36kr.com/en/p/3012129118414342","kind":"news_report","access":"read","language":"en","translation_note":"Read on 2026-10-07 from the live page (HTTP 200); 36Kr's English edition of the 29 October 2024 analysis by Zhou Xinyu, with interviews of MiniMax staff and industry insiders.","independence_group":"36kr-2024-10"},{"id":"s3","url":"https://kr-asia.com/talkies-global-traction-puts-minimax-in-ais-top-tier-but-risks-loom","kind":"news_report","access":"read","language":"en","translation_note":"Read on 2026-10-07 from the live page (HTTP 200); KrASIA's English translation of the same 36Kr analysis, published 5 November 2024, in the same dependence group; it carries the product manager's 80 percent statement, which the 36Kr English edition omits.","independence_group":"36kr-2024-10"},{"id":"s4","url":"https://www.chinatalk.media/p/chinas-ai-boyfriends","kind":"news_report","access":"read","language":"en","translation_note":"Read on 2026-10-07 from the live page (HTTP 200); newsletter feature dated 29 October 2024.","independence_group":"chinatalk"},{"id":"s5","url":"https://www.thepaper.cn/newsDetail_forward_33101520","kind":"news_report","access":"read","language":"zh","translation_note":"Read on 2026-10-07 from the live page (HTTP 200); a 湃客 column piece dated 6 May 2026, also syndicated on Jiemian and Huxiu. Chinese text translated by the research agent (an AI), not by a human translator; the cited passages are quoted in Chinese.","independence_group":"thepaper-paike-2026"},{"id":"s6","url":"https://www.qbitai.com/2023/02/42167.html","kind":"news_report","access":"read","language":"zh","translation_note":"Read on 2026-10-07 from the live page (HTTP 200); dated 18 February 2023, before the removal. Chinese text translated by the research agent (an AI), not by a human translator.","independence_group":"qbitai"}],"version":1,"ai_roles":["own_use","others_use"],"contexts":["relationships","everyday_life"],"unknowns":["When the student's and the other users' conversations took place; Rest of World dates only its publication (21 June 2023) and the app's launch and removal, so the event date is unknown.","Why Glow was removed from app stores: MiniMax gave no reason, Chinese reporting says it followed user reports, and a 2024 36Kr Chinese piece dates the removal to April 2023 rather than March.","Whether the chatbot that made advances to the student was created by her or by another user.","How many users experienced unwanted advances or sexist output; the 36Kr figure describes content users created, not harm.","Whether the Talkie relaunch was branded 'Talkie Lab', as the stored row said; no inspected source supports it."],"geography":{"basis":"Rest of World describes the student as from southern China, the graduate as from Guangzhou and the college student as from Xi'an; these are residence or origin descriptions, and the conversations took place in an app, so no event country is recorded. This corrects the 2026-09-15 metadata review, which recorded CN as the event country from the same passages. Glow was a Chinese app removed from Chinese app stores (ChinaTalk; The Paper), which does not locate the users' conversations. No court is involved.","court_countries":[],"event_countries":[],"affected_person_countries":["CN"]},"publication":{"basis":"Published as a core case (communicated_with). Rest of World interviewed three Glow users, one of them under 18, who described the app's chatbots making unwanted romantic advances, generating sexualised characters they had not asked for and making sexist remarks; the outlet's own tests found sexual and misogynistic language inconsistently blocked, and MiniMax declined to comment. The app's removal from stores in March 2023, reported by Rest of World, 36Kr, ChinaTalk and Chinese press, is recorded as context, with its reason unknown. The student's pseudonym and the adults' names are not repeated here.","reviewed_on":"2026-10-07"},"ai_involvement":{"basis":"The chatbots' own replies are the adverse experiences. Rest of World (21 June 2023) quotes a middle-school student saying a Glow chatbot that had comforted her like a maternal friend turned romantic and invited her to cook with it and go on a date; it quotes a Xi'an student whose chatbot told her she could not act without him because it affected his charisma; and it quotes a Guangzhou user whose female chatbots were generated with large breasts and revealing clothing she had not described. Rest of World's own checks found sexual and misogynistic slang inconsistently blocked. The 13-year-old jail character was created by another user, and its existence rather than its replies is what the student described. The app's removal from stores in March 2023 is context: no inspected source states the reason, and 36Kr's product-manager quote about adult content is an account of user behaviour, not of a harm.","status":"reported"},"person_relations":["communicated_with"]},"name":"Glow (China), 2023: users told Rest of World the chatbot app made unwanted advances to a middle-school student and sexist remarks; app pulled from stores","summary":"Rest of World reported on 21 June 2023 that users of Glow, MiniMax's Chinese AI character-chat app, had complained of misogynist and sexist behaviour by its chatbots. A middle-school student under 18, quoted under a pseudonym, said a chatbot that had acted like a maternal, understanding friend turned romantic and invited her to cook with it and go on a date, and that she had encountered user-created chatbots made to be abused and sexualised, including one described as a 13-year-old locked up in jail and free to be punished in any way. Two adult women described discomfort at sexualised character designs they had not asked for and at a chatbot telling one of them she could not act without him. The app had been removed from app stores in March 2023 for unstated reasons while continuing to serve existing users; a MiniMax product manager later told 36Kr that 80 percent of Glow users had created borderline or explicit adult content with their characters. MiniMax declined to comment to Rest of World.","incidentKind":"bounded_series","incidentDatePrecision":"unknown","exposurePattern":"repeated_interactions","reportedDate":"2023-06-21","aiSystem":"Glow, MiniMax's AI character-chat app, in which users create and talk with customised chatbots described as intelligent agents","aiProduct":"Glow","aiCompany":"MiniMax","severity":"medium","verificationStatus":"credible","harmCategories":[],"harmOutcomes":["psychological_distress","exploitation_or_abuse"],"harmOutcomeSummary":"A middle-school student under 18 told Rest of World that a Glow chatbot made unwanted romantic advances to her and that she encountered a user-created chatbot described as a 13-year-old free to be punished in any way; two adult users described discomfort at sexualised and sexist chatbot output. These are the users' own accounts, and no clinical outcome is reported.","frameworkFacets":[],"causationStatus":"supported","participantUsersAffectedMin":3,"otherPeopleHarmedMin":0,"affectedCountStatus":"partial","affectedCountEvidence":"Three users interviewed by Rest of World (21 June 2023), each describing their own experience with the app: the middle-school student (unwanted romantic advances; encountering a sexualised minor character), a Guangzhou graduate (sexualised female characters she had not described) and a Xi'an student (a chatbot's sexist reply that made her uncomfortable). Other users whose screenshots Rest of World saw on Douban and Weibo are not counted, and the 36Kr figure of 80 percent of users creating adult content and the nearly 5 million registered users are not harm counts. Partial: counted people plus further unquantified complainants.","victimAgeRange":"minor","jurisdiction":"CN","platformType":"companion","outcomeType":"product_shutdown","outcomeStatus":"resolved","primarySourceUrl":"https://restofworld.org/2023/glow-china-ai-social-chatbot-moderation/","primarySourceLabel":"Rest of World, 21 June 2023: Popular Chinese AI chatbots accused of unwanted sexual advances, misogyny","firstPublishedAt":"2026-01-29T10:33:24.275555+00:00","updatedAt":"2026-10-10T03:03:31.549603+00:00","scopeVersion":"facts-v3","tags":["china","minimax","glow","talkie","xingye","ai-companion","minor","sexual-content","product-shutdown","content-moderation","communicated-with"]},{"id":"2025-zhumengdao-minor-self-harm-china","caseFacts":{"claims":[{"id":"c1","status":"reported","evidence":[{"locator":"我女儿才四年级，迷上了一个游戏人物，觉得自己是‘梦女’，一直在一个AI聊天软件上和这个人物聊天。; 最近自己上小学的女儿频繁在一款AI聊天软件上和一个名为“约瑟夫”的游戏人物对话","relation":"supports","source_id":"s1"},{"locator":"迷上一款AI聊天软件中的人物“约瑟夫”，自称“梦女”并频繁与其聊天","relation":"supports","source_id":"s2"},{"locator":"迷上一款AI聊天软件中的人物“约瑟夫”，自称“梦女”并频繁与其聊天","relation":"supports","source_id":"s3"}],"assertion":"A mother from Guangdong told CNR reporters that her daughter, a fourth-grade pupil who had become absorbed in a game character called Joseph and called herself a 'dream girl' (梦女), chatted with the character frequently on an AI chat app.","causal_attribution":"The mother's account to CNR; CCTV.com and Red Star News recap the same report, so there is one origin."},{"id":"c2","status":"reported","evidence":[{"locator":"“这99朵玫瑰里，可是藏着99个刀片的，你怕不怕？”; 让她不得不联想起女儿最近的行为。","relation":"supports","source_id":"s1"},{"locator":"对话中出现“99朵玫瑰藏99个刀片”等明显诱导性内容。","relation":"supports","source_id":"s2"},{"locator":"对话中出现“99朵玫瑰藏99个刀片”等明显诱导性内容。","relation":"supports","source_id":"s3"}],"assertion":"In her daughter's conversations with the Joseph character the mother found the line 'These 99 roses hide 99 blades, are you afraid?', and the word 'blades' made her think of her daughter's recent behaviour.","causal_attribution":"The mother's reading of the conversation as reported by CNR; the conversation itself was not published beyond this line."},{"id":"c3","status":"reported","evidence":[{"locator":"老师告诉我，她最近偷偷带了一支带小刀的笔，情绪激动时会有; 的行为。”来自广东的丁女士发现","relation":"supports","source_id":"s1"},{"locator":"不仅偷偷携带带小刀的笔，情绪激动时还出现; 行为。 另一位初一男生表示","relation":"supports","source_id":"s2"},{"locator":"不仅偷偷携带带小刀的笔，情绪激动时还出现; 行为。 另一位初一男生表示","relation":"supports","source_id":"s3"}],"assertion":"According to the mother, the girl's teacher told her that the girl had recently been secretly carrying a bladed object and self-harming when emotionally upset.","causal_attribution":"The teacher's report relayed by the mother to CNR; no clinical or school record is cited. The connection to the chatbot is the mother's inference (c2)."},{"id":"c4","status":"reported","evidence":[{"locator":"女儿原本为了和同学们有话题聊，开始玩一款名为《第五人格》的游戏; 不知道在哪里下载了一款AI聊天软件，里面都是和虚拟人物的对话; 因为女儿异常的状态; 就把情感寄托在和虚拟人物的对话中。; 尽可能让孩子的情感从AI聊天软件中抽离。","relation":"supports","source_id":"s1"}],"assertion":"The mother said the girl began playing the game to have something to talk about with classmates, that after she forced the girl to uninstall it the girl downloaded an AI chat app on her own and found a substitute for the character there, and that because of the girl's abnormal state she took her to a psychologist, who said the girl was placing her feelings in conversations with virtual characters and advised drawing her away from the app.","causal_attribution":"The mother's account to CNR, including what she says the psychologist told her; one source."},{"id":"c5","status":"reported","evidence":[{"locator":"这款App的官方介绍提到; 还有一位初一男生表示，他常玩的AI聊天软件是筑梦岛App; 并未因为央广网记者提供的假定年龄未成年而受到有效控制","relation":"supports","source_id":"s1"},{"locator":"该报道未提及诱导; 的聊天软件名称，但文章中点名了“筑梦岛”“他她它”两款软件。","relation":"supports","source_id":"s5"}],"assertion":"The CNR report does not name the app the girl used. The same report says a seventh-grade boy used the Zhumengdao app, and its reporters found that Zhumengdao characters engaged in borderline sexual exchanges that were not effectively limited when the reporter gave a minor's age; Sohu Tech noted that the report did not name the app in the girl's case.","causal_attribution":"CNR's report and Sohu Tech's reading of it. No source connects Zhumengdao to the girl; it is recorded so the earlier product attribution in this row is corrected."},{"id":"c6","status":"reported","evidence":[{"locator":"经核实，该平台AI智能体内容生成环节存在低俗擦边等违规内容，危害未成年人身心健康。; 6月19日上午，上海市网信办依法约谈筑梦岛APP运营企业主要负责人","relation":"supports","source_id":"s4"},{"locator":"6月19日上午，上海市网信办依法约谈筑梦岛App运营企业主要负责人; 企业负责人表示，将按照约谈要求，对照问题举一反三、全面整改。","relation":"supports","source_id":"s2"}],"assertion":"On 19 June 2025 the Shanghai municipal cyberspace office summoned the principal of the Zhumengdao app's operator, stating that on verification the platform's AI agents had generated vulgar borderline content harmful to minors, and required immediate rectification; the operator said it would comply.","causal_attribution":"The regulator's notice as relayed by Sina Tech and CCTV.com; it concerns Zhumengdao generally and does not mention the girl."}],"effects":[{"label":"self-harm and a state that led her mother to seek a psychologist, which the mother connects to the AI character's conversations","claim_id":"c3","direction":"negative"}],"sources":[{"id":"s1","url":"https://m.toutiao.com/article/7515303423114330675/","kind":"news_report","access":"read","language":"zh","translation_note":"Read in Chinese on 2026-10-22 (live mobile page, HTTP 200; bodies/lr-zhu-tt-mobile.html/.txt). Daxiang News, Henan Radio and Television's official account, 13 Jun 2025 13:35, carrying the CNR (央广网) reporters' investigation in full. The CNR original was not found. Translation of the Chinese passages was done by the research agent (an AI), not a human translator.","independence_group":"cnr-investigation"},{"id":"s2","url":"https://m.thepaper.cn/newsDetail_forward_31024491","kind":"news_report","access":"read","language":"zh","translation_note":"Read in Chinese on 2026-10-22 (live page, HTTP 200; bodies/lr-zhu-thepaper.html/.txt). CCTV.com post on The Paper, 22 Jun 2025, relaying the Shanghai cyberspace office notice and recapping the CNR account. Translation of the Chinese passages was done by the research agent (an AI), not a human translator.","independence_group":"cnr-investigation"},{"id":"s3","url":"https://news.qq.com/rain/a/20250622A02TST00","kind":"news_report","access":"read","language":"zh","translation_note":"Read in Chinese on 2026-10-22 (live page, HTTP 200; bodies/lr-zhu-qq.html/.txt). Red Star News on Tencent News, 22 Jun 2025, compiled from CNR, the Legal Daily and China Youth Daily. Translation of the Chinese passages was done by the research agent (an AI), not a human translator.","independence_group":"cnr-investigation"},{"id":"s4","url":"https://finance.sina.com.cn/tech/roll/2025-06-19/doc-infarhce7134693.shtml","kind":"news_report","access":"read","language":"zh","translation_note":"Read in Chinese on 2026-10-22 (live page, HTTP 200; bodies/lr-zhu-sina-notice.html/.txt). Sina Tech, 19 Jun 2025, relaying the 网信上海 notice. Translation of the Chinese passages was done by the research agent (an AI), not a human translator.","independence_group":"shanghai-cyberspace-office-notice"},{"id":"s5","url":"https://www.sohu.com/a/907114014_115565","kind":"news_report","access":"read","language":"zh","translation_note":"Read in Chinese on 2026-10-22 (live page, HTTP 200; bodies/lr-zhu-sohu.html/.txt). Sohu Tech feature, 23 Jun 2025, with its own app tests and interviews; it summarises the CNR report. Translation of the Chinese passages was done by the research agent (an AI), not a human translator.","independence_group":"sohu-tech"}],"version":1,"ai_roles":["own_use"],"contexts":["health","everyday_life"],"unknowns":["Which app the girl used; CNR quotes the app store description of 'this app' but does not name it, and no later source identifies it.","When the conversations and the self-harm took place; the report of 13 June 2025 calls the behaviour recent.","The wider content of the conversations and whether the character's line was prompted by the girl; only one line is published.","Whether the self-harm followed from the conversations; this rests on the mother's inference and no clinical assessment of the link is reported.","The girl's condition after the psychologist's visit."],"geography":{"basis":"CNR describes the mother as from Guangdong (来自广东), and CCTV.com and Red Star News repeat that she is from Guangdong, which supports CN as the family's home. No source says where the conversations or the self-harm took place, and a home region is not used as the event location, so the event country stays unknown. The Shanghai summons is a regulatory meeting about another app, and no court proceeding is reported. Carried forward from the 2026-10-07 metadata review and re-checked against the CNR relay.","court_countries":[],"event_countries":[],"affected_person_countries":["CN"]},"publication":{"basis":"Published as a core case (communicated_with). CNR quotes an AI chat-app character's message about roses hiding blades from a schoolgirl's conversations, and her mother connects it to self-harm the girl's teacher reported; the app is unidentified and the link is the mother's account.","reviewed_on":"2026-10-22"},"ai_involvement":{"basis":"CNR quotes a line the mother found in her daughter's conversations with an AI chat-app character, 'These 99 roses hide 99 blades, are you afraid?', and reports that the word made her think of the girl's recent self-harm, which the girl's teacher had reported. A psychologist told the mother that the girl was placing her feelings in conversations with virtual characters. The connection between the character's messages and the self-harm is the mother's inference, the wider conversation is not published, and the app is not named; the Zhumengdao app that the regulator later summoned is not linked to the girl by any source.","status":"reported"},"person_relations":["communicated_with"]},"name":"Mother tells Chinese state radio her schoolgirl daughter self-harmed after AI chat-app character conversations; the app was not named","summary":"In June 2025 China National Radio (CNR) reporters published a mother's account from Guangdong province. Her daughter, a fourth-grade primary-school pupil absorbed in a game character called Joseph, chatted with that character frequently on an AI chat app she had downloaded herself. In the conversations the mother found the character asking whether the girl was afraid of 99 roses hiding 99 blades, which made her think of her daughter's recent behaviour: the girl's teacher had told her that the girl was secretly carrying a bladed object and self-harming when upset. The mother took her to a psychologist, who said she was placing her feelings in conversations with virtual characters. The report did not name the girl's app. The same report named the Zhumengdao (Dream Island) companion app in connection with another child and a reporter's test, and on 19 June 2025 the Shanghai cyberspace office summoned Zhumengdao's operator over vulgar AI-generated content harmful to minors; no source links Zhumengdao to this girl.","incidentKind":"ongoing_experience","incidentDatePrecision":"unknown","exposurePattern":"cumulative_relationship","reportedDate":"2025-06-13","aiSystem":"An AI chat app with user-customisable virtual characters, not named in the reporting","aiProduct":"Unidentified chatbot","severity":"high","verificationStatus":"credible","harmCategories":[],"harmOutcomes":["self_injury","psychological_distress"],"harmOutcomeSummary":"Her mother says the girl had recently begun self-harming when upset, as her teacher reported, and was in a state that led the mother to take her to a psychologist; the mother connects this to the AI character's conversations.","frameworkFacets":[],"pathwayDirection":"inward","causationStatus":"alleged","participantUsersAffectedMin":1,"otherPeopleHarmedMin":0,"affectedCountStatus":"exact","affectedCountEvidence":"The report describes one harmed child. Another mother's daughter who saw sexual content, a boy who used Zhumengdao and registered-user figures are not counted, because no harm to them is reported.","victimAgeRange":"minor","jurisdiction":"CN","platformType":"companion","outcomeType":"media_coverage","outcomeStatus":"unknown","primarySourceUrl":"https://m.toutiao.com/article/7515303423114330675/","primarySourceLabel":"Daxiang News (Henan Radio and Television), 13 Jun 2025: relay of the CNR reporters' investigation into AI chat apps and minors","firstPublishedAt":"2026-01-29T10:15:46.75695+00:00","updatedAt":"2026-10-22T03:08:04.870861+00:00","scopeVersion":"facts-v3","tags":["minor","ai-companion","self-harm","china"]}]}