{"meta":{"exportedAt":"2026-11-20T00:01:39.090Z","formatVersion":2,"selection":{"q":"hiring","system":"","harm":"","context":"","country":"","role":"","relation":"","evidence":"","year":"2026","response":"","severity":"","verification":"","view":"incidents","sort":"added"},"totalIncidents":2,"coverage":{"cases":2,"countries":2,"languages":2,"unknownLocation":0,"locationPending":0,"unknownLanguage":0,"unknownDate":0,"lawsuits":0,"regulatory":0,"minors":0,"coreRelations":1,"contextualRelations":1,"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-san-francisco-andon-labs-ai-store-manager-agent-recommended-dismissing-an-employee-staff-carried-it-out","caseFacts":{"claims":[{"id":"c1","status":"reported","evidence":[{"locator":"'Since April, an AI agent named Luna has been running Andon Market, a store in San Francisco.'; 'everyone working at Andon Market is formally employed by Andon Labs, with guaranteed pay, fair wages, and full legal protections'; 'This is a controlled experiment'","relation":"supports","source_id":"s1"},{"locator":"'Luna, built using Anthropic's Claude models, selected merchandise, hired contractors, posted jobs on Indeed, interviewed applicants, and hired employees for Andon Market'","relation":"supports","source_id":"s3"},{"locator":"'the workers hired by Claude to work in the San Francisco store are real people with genuine employment contracts'","relation":"supports","source_id":"s2"}],"assertion":"Andon Market in San Francisco is managed by an AI agent, Luna, built on Anthropic's Claude models; Andon Labs calls the store a controlled experiment, and its workers are employed by Andon Labs under genuine employment contracts.","causal_attribution":"Establishes the setting and the AI system; all from the operator's account."},{"id":"c2","status":"reported","evidence":[{"locator":"'Time passed and the employee handbook disappeared from Luna’s memory. During this time, the new employee was late over and over again.'","relation":"supports","source_id":"s1"},{"locator":"'The worker in question was fired for being late on 17 out of their 23 shifts, according to Andon Labs.'","relation":"supports","source_id":"s2"}],"assertion":"The employee was repeatedly late after the handbook Luna had written dropped out of its memory; TIME, citing Andon Labs, reports lateness on 17 of 23 shifts.","causal_attribution":"Background to the recommendation; the figures are the operator's."},{"id":"c3","status":"reported","evidence":[{"locator":"'Luna found her policies and came to the conclusion that it was time for a verbal warning.'; 'I want you to think about if this is really the right fit.'; 'Recommendation: I lean toward parting ways'","relation":"supports","source_id":"s1"},{"locator":"'Even then, Claude’s first recommendation was to formally warn this employee, not to fire them.'; 'Petersson acknowledges was “a leading question”'","relation":"supports","source_id":"s2"}],"assertion":"Asked to review its policies, Luna first proposed a verbal warning; after a staff member said formal conversations had taken place and asked it to consider whether this was the right fit, it wrote that it leaned toward parting ways.","causal_attribution":"The AI output that informed the decision, with the human prompt that preceded it; one origin (the company's logs)."},{"id":"c4","status":"reported","evidence":[{"locator":"'the termination itself was reviewed and delivered by humans'","relation":"supports","source_id":"s1"},{"locator":"'a decision that the humans in the lab reviewed and carried out.'","relation":"supports","source_id":"s3"},{"locator":"'Andon Labs announced Friday that the AI boss, Luna, had fired an unnamed employee'; 'Axel Backlund of Andon Labs said high-stakes decisions always have human oversight'","relation":"supports","source_id":"s4"}],"assertion":"The employee was dismissed; Andon Labs says the termination was reviewed and delivered by its staff.","causal_attribution":"Records the adverse decision and the human role in carrying it out."},{"id":"c5","status":"reported","evidence":[{"locator":"'TIME reached out to the fired worker for comment, but did not receive a reply'","relation":"supports","source_id":"s2"},{"locator":"'In this instance, we did not think that that was necessary because the firing was warranted,'","relation":"supports","source_id":"s3"}],"assertion":"Andon Labs says the firing was warranted, and the dismissed employee did not reply to TIME's request for comment.","causal_attribution":"Counterevidence and the absence of the employee's own account."}],"effects":[{"label":"Employee dismissed after the AI manager agent recommended parting ways","claim_id":"c4","direction":"negative"}],"sources":[{"id":"s1","url":"https://andonlabs.com/blog/ai-bosses-2","kind":"organization_report","access":"read","language":"en","translation_note":"Read in English on 2026-10-28 (HTTP 200, full post with transcript appendix). The operator's own account and logs; employee messages are paraphrased by the company.","independence_group":"andon-labs-blog-2026-08-14"},{"id":"s2","url":"https://time.com/article/2026/08/14/claude-fired-worker-ai-job-disruption/","kind":"news_report","access":"read","language":"en","translation_note":"Read in English on 2026-10-28 through an Internet Archive capture (live page returned HTTP 406). TIME saw the management logs and interviewed Andon Labs' chief executive and a remaining employee.","independence_group":"time-2026-08-14"},{"id":"s3","url":"https://www.businessinsider.com/ai-running-sf-store-fired-employee-for-the-first-time-2026-8","kind":"news_report","access":"read","language":"en","translation_note":"Read in English on 2026-10-28 (HTTP 200). Relies on Andon Labs' report, logs and an interview with its cofounder.","independence_group":"business-insider-2026-08-15"},{"id":"s4","url":"https://sfstandard.com/2026/08/17/ai-boss-fires-worker/","kind":"news_report","access":"read","language":"en","translation_note":"Read in English on 2026-10-28 (HTTP 200). Relies on Andon Labs' announcement and an interview with a remaining employee.","independence_group":"sf-standard-2026-08-17"}],"version":1,"ai_roles":["institutional_use"],"contexts":["work"],"unknowns":["The exact date of the dismissal: the company's logs place the recommendation on 16 June 2026 and plan the dismissal before 28 June, while TIME's 14 August report says the firing happened the previous month.","The dismissed employee's own account and whether they contested the dismissal.","Whether the staff would have dismissed the employee without the agent's recommendation."],"geography":{"basis":"The store where the employee worked is Andon Market in San Francisco (Andon Labs blog, TIME, San Francisco Standard). The employee's own country is not stated.","court_countries":[],"event_countries":["US"],"affected_person_countries":[]},"publication":{"basis":"Published because the operator's own logs show its AI manager agent recommending that an employee be dismissed and the company says its staff then carried out the dismissal. The account is the company's, the human prompt that preceded the recommendation is recorded, and the employee's own view is unknown.","reviewed_on":"2026-10-28"},"ai_involvement":{"basis":"Andon Labs published the agent's own messages: after reviewing the employee's record it wrote 'I lean toward parting ways' and proposed how to carry out the dismissal. The company says its staff then reviewed and delivered the termination, so the agent's recommendation informed a human decision to dismiss the employee. The logs also show the agent first proposed a verbal warning and reached its recommendation after a staff member's prompt that the company's chief executive called a leading question.","status":"reported"},"person_relations":["informed_decision_about"]},"name":"San Francisco: Andon Labs says its AI store-manager agent recommended firing an employee over repeated lateness, and its staff carried out the dismissal","summary":"Andon Labs, an AI research company, runs Andon Market, a store in San Francisco managed by an AI agent called Luna built on Anthropic's Claude models. The company calls the store a controlled experiment; its workers are formally employed by Andon Labs, and TIME describes them as real people with genuine employment contracts. In a blog post of 14 August 2026 the company said Luna had decided to fire one of its employees. According to the logs it published, an employee handbook Luna had written dropped out of its memory and it did not act on the employee's repeated lateness; when Andon Labs asked it to search its memory, it first proposed a verbal warning, and after a staff member said formal conversations had already taken place and asked it to think about whether this was the right fit, it wrote that it leaned toward parting ways. Andon Labs says the termination was reviewed and delivered by humans. TIME, citing Andon Labs, reports that the employee was late for 17 of 23 shifts; the company says the firing was warranted, and the employee did not reply to TIME.","incidentDate":"2026-06-01","incidentKind":"single_event","incidentDatePrecision":"month","exposurePattern":"repeated_interactions","reportedDate":"2026-08-14","aiSystem":"Luna, Andon Labs' AI agent that manages the Andon Market store (hiring, schedules, time off and staff messages), running on Anthropic's Claude Opus 4.8 when it recommended the dismissal, according to Andon Labs","aiProduct":"Andon Labs Luna agent","severity":"low","verificationStatus":"credible","harmCategories":[],"harmOutcomes":["professional_harm"],"harmOutcomeSummary":"An employee lost their job after the store's AI manager agent recommended parting ways, according to Andon Labs, which says its staff reviewed and delivered the termination.","frameworkFacets":[],"causationStatus":"supported","participantUsersAffectedMin":0,"otherPeopleHarmedMin":1,"affectedCountStatus":"exact","affectedCountEvidence":"One dismissed employee, described by Andon Labs, TIME and Business Insider. Remaining staff are not counted.","victimAgeRange":"unknown","jurisdiction":"US","platformType":"agent","outcomeType":"internal_action","outcomeStatus":"resolved","primarySourceUrl":"https://andonlabs.com/blog/ai-bosses-2","primarySourceLabel":"Andon Labs blog, 14 Aug 2026: AI bosses are slow to fire and quick to hire","firstPublishedAt":"2026-10-28T03:07:59.840936+00:00","updatedAt":"2026-10-28T03:07:59.840936+00:00","scopeVersion":"facts-v3","tags":["work","dismissal","ai-agent","ai-manager","institutional-use"]},{"id":"2026-japan-it-firm-online-job-interview-applicant-likely-used-ai-deepfake-to-impersonate-real-engineer","caseFacts":{"claims":[{"id":"c1","status":"reported","evidence":[{"locator":"今月上旬、東京都内のＩＴ企業が行った中途採用のオンライン面接。 ... ２分ほどで面接は終了した。","relation":"supports","source_id":"s1"}],"assertion":"In early March 2026 a Tokyo IT company held an online mid-career interview with an applicant who asked for fully remote work abroad and ended the interview after about two minutes when told the job required office attendance.","causal_attribution":"Yomiuri, citing the company."},{"id":"c2","status":"reported","evidence":[{"locator":"の学歴や経歴は、男性と同じだった。","relation":"supports","source_id":"s1"},{"locator":"AIの映像を使って「実在する別人になりすました応募者」が現れた。 ... 「面談を受けていない」と回答があり、面談相手がなりすましだったことが確認された。","relation":"supports","source_id":"s2"}],"assertion":"The applicant's CV and online appearance matched a real engineer, who had not applied for or attended the interview.","causal_attribution":"Yomiuri and Business Insider Japan, from the company's and the engineer's accounts."},{"id":"c3","status":"reported","evidence":[{"locator":"ディープフェイク判定技術を持つ東大発ベンチャー「ナブラス」（東京）、米ＩＴ大手「オクタ」など三つの企業・研究機関に調べてもらった。その結果、「ディープフェイクである可能性が高い」「生成ＡＩで作られた可能性がある」「ほとんどフェイク」との判定だった。 ... 目の位置がずれる瞬間がある▽口の動きと音声が合っていない","relation":"supports","source_id":"s1"}],"assertion":"Three companies and research bodies that Yomiuri asked to examine the interview video, including Nabras and Okta, judged it likely a deepfake, possibly generative-AI made, or almost certainly fake, citing boundary, eye-position and lip-sync anomalies.","causal_attribution":"Yomiuri's commissioned examinations."},{"id":"c4","status":"reported","evidence":[{"locator":"送られてきた動画を見て「ネットに公開されている自分の顔写真や動画を使い、ＡＩでなりすまされたのかも」と思い、怖くなった。 ... 「自分になりすました人物が採用されているかもしれず、恐ろしい」と話し","relation":"supports","source_id":"s1"}],"assertion":"After viewing the video, the engineer thought their public photos and videos had been used to impersonate them with AI, was frightened, and said they feared someone impersonating them might be hired.","causal_attribution":"The engineer's statements to Yomiuri."},{"id":"c5","status":"reported","evidence":[{"locator":"オンライン会議ツール上で表示されていた映像の解像度は極端に低く、顔の動きや歯の見え方にも不自然さがあったという。 ... 合成音声ではなく、ある程度日本語を理解する人物の肉声だった可能性が高い","relation":"supports","source_id":"s2"}],"assertion":"The interviewing CTO described an extremely low-resolution picture with unnatural face and teeth movement, and judged the voice probably a real person's rather than synthetic.","causal_attribution":"The CTO's account to Business Insider Japan."},{"id":"c6","status":"reported","evidence":[{"locator":"都内の別の企業からも「応募があった」と連絡があったという。","relation":"supports","source_id":"s1"}],"assertion":"Another Tokyo company told the engineer it had received an application, after the engineer asked on X for similar cases.","causal_attribution":"Yomiuri, citing the engineer."}],"effects":[{"label":"a real engineer's name, career and apparent face were used by a job applicant, and the engineer said they were frightened that an impersonator might be hired","claim_id":"c4","direction":"negative"}],"sources":[{"id":"s1","url":"https://www.yomiuri.co.jp/national/20260319-GYT1T00020/","kind":"news","access":"read","language":"ja","translation_note":"Read in Japanese by the research agent (an AI) on 2026-10-19 and translated by it; no human translator or reviewer was involved. Yomiuri Shimbun, 19 March 2026, by-lined report including its commissioned video examinations. The examiners' reports themselves, the interview video and the X posts were not inspected.","independence_group":"yomiuri-deepfake-interview-report"},{"id":"s2","url":"https://www.businessinsider.jp/article/2603-deepfake-job-interview-risk-japan/","kind":"news","access":"read","language":"ja","translation_note":"Read in Japanese by the research agent (an AI) on 2026-10-19 (pages 1 and 2) and translated by it; no human translator or reviewer was involved. Business Insider Japan, 19 March 2026, based on its own interviews with the interviewing company's CTO and the impersonated engineer.","independence_group":"business-insider-japan-interviews"}],"version":1,"ai_roles":["others_use"],"contexts":["work","privacy"],"unknowns":["Who the applicant was and what the applicant intended.","Which tool produced the altered video.","Whether the two outlets describe the same company's interview.","Whether any police report was made."],"geography":{"basis":"Yomiuri reports the interview was held by a Tokyo IT company and that the impersonated engineer heads a Tokyo software company. The applicant's location is unknown. No court proceeding is reported.","court_countries":[],"event_countries":["JP"],"affected_person_countries":["JP"]},"publication":{"basis":"Published as a core case (depicted_or_impersonated): deepfake examiners commissioned by Yomiuri judged that the video of a job applicant who used a real engineer's name and career was likely AI-made, and the engineer said the impersonation was frightening. The AI involvement is suspected on the strength of that examination; the applicant is unidentified.","reviewed_on":"2026-10-19"},"ai_involvement":{"basis":"Yomiuri had the interview video examined by three companies and research bodies, including the deepfake-detection venture Nabras and the US IT company Okta; they judged it likely a deepfake, possibly generative-AI made, or almost certainly fake, and listed boundary, eye-position and lip-sync anomalies. That examined video is the one that showed the applicant with the engineer's apparent face. The examiners' wording is probabilistic, the tool and operator are unidentified, and the interviewing CTO thought the voice was a real person's.","status":"suspected"},"person_relations":["depicted_or_impersonated"]},"name":"Japan: deepfake checks commissioned by Yomiuri find an online job applicant likely used AI to impersonate a real engineer's face and identity","summary":"The Yomiuri Shimbun reported on 19 March 2026 that in early March an applicant in a Tokyo IT company's online mid-career job interview used the name, career history and apparent face of a real engineer who heads a Tokyo software company. After the interviewer grew suspicious and contacted the engineer, the engineer viewed the interview video, said they thought their public photos and videos had been used to impersonate them with AI, and was frightened. Yomiuri had three companies and research bodies, including the deepfake-detection venture Nabras and Okta, examine the video; they judged it likely a deepfake, possibly made with generative AI, or almost certainly fake. Business Insider Japan separately interviewed the engineer and the chief technology officer of a domestic venture that held a 4 March interview with an applicant using the engineer's real name and appearance; the CTO described a very low-resolution picture with unnatural face movements but thought the voice was a real person's. Who the applicant was is not established.","incidentDate":"2026-03-01","incidentKind":"single_event","incidentDatePrecision":"month","exposurePattern":"single_interaction","reportedDate":"2026-03-19","aiSystem":"An unidentified AI face-altering video tool, which three examiners commissioned by Yomiuri judged had likely been used to make the applicant's interview video (tool and operator not identified)","aiProduct":"Unidentified video tool (suspected)","severity":"low","verificationStatus":"credible","harmCategories":[],"harmOutcomes":["psychological_distress"],"harmOutcomeSummary":"According to the Yomiuri Shimbun and Business Insider Japan, an applicant in a company's online job interview used a real engineer's name, career and apparent face, and the engineer said they were frightened that someone impersonating them might be hired.","frameworkFacets":[],"causationStatus":"alleged","participantUsersAffectedMin":0,"otherPeopleHarmedMin":1,"affectedCountStatus":"documented_minimum","affectedCountEvidence":"One person is counted: the engineer whose identity and apparent face were used and who said they were frightened. The interviewers and the other company that received an application under the engineer's name are not counted as harmed people, so the figure is a lower bound.","victimAgeRange":"adult","jurisdiction":"JP","platformType":"other","outcomeType":"media_coverage","outcomeStatus":"unknown","primarySourceUrl":"https://www.yomiuri.co.jp/national/20260319-GYT1T00020/","primarySourceLabel":"Yomiuri Shimbun, 19 March 2026: online job interview applicant impersonated a real person with generative AI","firstPublishedAt":"2026-10-19T03:06:40.830131+00:00","updatedAt":"2026-10-19T03:06:40.830131+00:00","scopeVersion":"facts-v3","tags":["deepfake","impersonation","job-interview","hiring","japan","depicted-or-impersonated"]}]}