{"license":"https://creativecommons.org/licenses/by/4.0/","attribution":"mikias.io citations board, https://mikias.io/citations/incidents","updatedAt":"2026-07-08","model":"Fully automated aggregator. The bulk layer is every documented court case from Damien Charlotin's AI Hallucination Cases database (CC0, credited, refreshed weekly from his published CSV into legal-cases.json). The cross-domain layer below covers incidents courts never see: each entry links the record of record (court or government document, the org's own statement or correction, or named-document press), is summarized in our own words with every characterization attributed, and is written by the weekly evidence sweep (content-engine/evidence.md) with deterministic validation. No opinion fields, no human gate. Spec: docs/career-pivot/fable-sprint/incidents-board/02-SPEC.md.","aggregateStats":[{"label":"PubMed-indexed papers with fabricated references","value":"1 in 277","asOf":"first 7 weeks of 2026","trend":"1 in 2,828 in 2023, then 1 in 458 in 2025","source":"https://www.thelancet.com/journals/lancet/article/PIIS0140-6736(26)00603-3/fulltext","refreshRule":"update when the CITADEL audit publishes a new window"},{"label":"Incidents logged by the AI Incident Database in 2025","value":"362","asOf":"2025","trend":"up from 233 in 2024","source":"https://hai.stanford.edu/ai-index/2026-ai-index-report/responsible-ai","refreshRule":"refresh yearly from the AI Index responsible-AI chapter"}],"incidents":[{"id":"kpmg-agentic-2026","date":"2026-06","org":"KPMG","jurisdiction":"Global (US-published)","artifact":"Total Experience agentic-AI thought-leadership report","category":"consulting-report","status":"withdrawn","sourceTier":"medium-high","failure":"GPTZero's review found only 5 of 45 citations accurate, with the rest fabricated, mangled, or misattributed, and the report was pulled in June 2026. The incident is logged by the OECD AI Incidents Monitor.","primaryUrl":"https://gptzero.me/news/investigations-kpmg/","sources":[{"url":"https://oecd.ai/en/incidents/2026-06-12-b883","type":"database-pointer","label":"OECD AIM"}],"addedAt":"2026-07-08","lastVerifiedAt":"2026-07-08","corrections":[],"orgResponse":null},{"id":"ey-canada-cyber-2026","date":"2026-05","org":"EY Canada","jurisdiction":"Canada","artifact":"Cybersecurity report on loyalty-program fraud, removed from the firm's website","category":"consulting-report","status":"withdrawn","sourceTier":"medium-high","failure":"GPTZero's forensic review found 16 of 27 cited sources fabricated, misattributed, or broken, and the report was pulled from EY's website in May 2026, according to trade press coverage.","primaryUrl":"https://gptzero.me/investigations/ey","sources":[{"url":"https://www.computing.co.uk/news/2026/ai/ey-cybersecurity-report-withdrawn-ai-hallucinations","type":"press","label":"Computing"},{"url":"https://www.lawyer-monthly.com/2026/05/ey-ai-report-fake-citations/","type":"press","label":"Lawyer Monthly"}],"addedAt":"2026-07-08","lastVerifiedAt":"2026-07-08","corrections":[],"orgResponse":null},{"id":"south-africa-ai-policy-2026","date":"2026-04","org":"Government of South Africa","jurisdiction":"South Africa","artifact":"Draft National Artificial Intelligence Policy","category":"policy-document","status":"withdrawn","sourceTier":"high","failure":"Communications Minister Solly Malatsi announced the draft policy's withdrawal on 26 April 2026 after at least 6 of its 67 bibliography sources were found to be fabricated, according to the government's own statement and independent reporting.","primaryUrl":"https://www.sanews.gov.za/south-africa/minister-announces-withdrawal-draft-ai-policy","sources":[{"url":"https://www.reuters.com/world/africa/south-africa-withdraws-ai-policy-due-fake-ai-generated-sources-2026-04-27/","type":"press","label":"Reuters"},{"url":"https://www.news24.com/business/tech/govts-draft-ai-policy-cites-fictitious-references-experts-believe-are-ai-hallucinations-20260424-1085","type":"press","label":"News24"},{"url":"https://incidentdatabase.ai/cite/1467/","type":"database-pointer","label":"AIID 1467"}],"addedAt":"2026-07-08","lastVerifiedAt":"2026-07-08","corrections":[{"date":"2026-07-08","note":"Earlier drafts of this dataset sourced this entry to a single aggregator newsletter. It is now sourced to the government's own withdrawal statement, with independent press and the AI Incident Database entry alongside."}],"orgResponse":null},{"id":"deloitte-au-dewr-2025","date":"2025-10","org":"Deloitte Australia","jurisdiction":"Australia","artifact":"AU$440K assurance review of the Targeted Compliance Framework for the Department of Employment and Workplace Relations","category":"consulting-report","status":"refunded","sourceTier":"high","failure":"The report contained fabricated academic references and a fabricated Federal Court quote, according to the department and press review. Deloitte repaid the final installment of its contract, and a disclosure of Azure OpenAI use was added only in the revised version.","primaryUrl":"https://www.dewr.gov.au/assuring-integrity-targeted-compliance-framework/resources/targeted-compliance-framework-assurance-review-final-report","sources":[{"url":"https://www.theguardian.com/australia-news/2025/oct/06/deloitte-to-pay-money-back-to-albanese-government-after-using-ai-in-440000-report","type":"press","label":"The Guardian"},{"url":"https://www.afr.com/companies/professional-services/deloitte-to-refund-government-after-admitting-ai-errors-in-440k-report-20251005-p5n05p","type":"press","label":"AFR"},{"url":"https://fortune.com/2025/10/07/deloitte-ai-australia-government-report-hallucinations-technology-290000-refund/","type":"press","label":"Fortune"},{"url":"https://incidentdatabase.ai/cite/1193/","type":"database-pointer","label":"AIID 1193"}],"addedAt":"2026-07-08","lastVerifiedAt":"2026-07-08","corrections":[{"date":"2026-07-08","note":"Earlier drafts of this dataset cited a refund of about AU$97K from press estimates. No primary source states a dollar amount, so the entry now says final installment only."}],"orgResponse":null},{"id":"deloitte-nl-health-2025","date":"2025-05","org":"Deloitte (Newfoundland and Labrador)","jurisdiction":"Canada (NL)","artifact":"About C$1.6M Health Human Resources Plan for the Government of Newfoundland and Labrador","category":"consulting-report","status":"acknowledged","sourceTier":"medium-high","failure":"Reporting in November 2025 identified at least four citations to non-existent sources in the plan delivered that May, and Deloitte acknowledged selective use of AI in its preparation, according to coverage by The Independent and Fortune.","primaryUrl":"https://theindependent.ca/news/lji/major-n-l-healthcare-report-contains-errors-likely-generated-by-a-i/","sources":[{"url":"https://www.gov.nl.ca/releases/2025/health/0529n03/","type":"primary","label":"gov.nl.ca release (the plan)"},{"url":"https://fortune.com/2025/11/25/deloitte-caught-fabricated-ai-generated-research-million-dollar-report-canada-government/","type":"press","label":"Fortune"},{"url":"https://incidentdatabase.ai/cite/1286/","type":"database-pointer","label":"AIID 1286"}],"addedAt":"2026-07-08","lastVerifiedAt":"2026-07-08","corrections":[{"date":"2026-07-08","note":"Earlier drafts of this dataset gave the cost as about $1.2M and dated the incident 2026. Independent coverage consistently reports C$1.6M, the plan was delivered in May 2025, and press coverage broke in November 2025."}],"orgResponse":null},{"id":"stanford-reglab-legal-tools-2024","date":"2024-05","org":"Commercial legal AI tools (Lexis+ AI, Westlaw AI-Assisted Research)","jurisdiction":"United States","artifact":"Stanford RegLab preregistered evaluation, Hallucination-Free?","category":"study","status":"published","sourceTier":"high","failure":"The preregistered evaluation found Lexis+ AI incorrect or misgrounded on more than 17% of queries and Westlaw's AI-Assisted Research on about 33%, while retrieval-augmented generation was being marketed as eliminating hallucinations.","primaryUrl":"https://reglab.stanford.edu/publications/hallucination-free-assessing-the-reliability-of-leading-ai-legal-research-tools/","sources":[{"url":"https://arxiv.org/pdf/2405.20362","type":"primary","label":"arXiv 2405.20362"},{"url":"https://hai.stanford.edu/news/ai-trial-legal-models-hallucinate-1-out-6-or-more-benchmarking-queries","type":"press","label":"Stanford HAI"}],"addedAt":"2026-07-08","lastVerifiedAt":"2026-07-08","corrections":[],"orgResponse":null}]}