7 incidents

Risk: Made-up output relied on Control: Drift monitoring Outcome: Not stated in sources

EffectiveNot effectiveNot effective (bypassed)Did not existNot stated in sourcesNot yet checkedAttacker's ownOrder: model · harness · application · target

Date	Incident	Risk	Controls mentioned in sources
2026-09-21	Frontier LLM agent fabricated a complete calibration-comparison report for commands that never ran	Made-up output relied on	EffectiveIndependent check of task results L3
2026-05-26	Pinsent Masons reprimanded over AI hallucinations in UK court filings	Made-up output relied on	Not effectiveHuman approval for high-impact actions L3 Did not existUser policy and training L3
2026-05-13	Canadian court annuls arbitral award over delegation to AI	Made-up output relied on	Did not existGrounding and fact checks L3
2026-03-20	Senior European journalist suspended over AI-generated quotes	Made-up output relied on	Did not existGrounding and fact checks L3
2026-02-09	Clinco v. Commissioner: legal hallucinations and evidentiary failures in AI-assisted filing	Made-up output relied on	Did not existGrounding and fact checks L3
2025-11-04	Deloitte report with AI-fabricated citations for an Australian federal welfare review (DEWR)	Made-up output relied on	Not effectiveUser policy and training L3
2025-07-01	Tennessee woman jailed six months after AI tool misidentified her	Made-up output relied on	EffectiveGrounding and fact checks L3