AI Incidents
Real-world cases where AI systems caused harm, classified by risk domain and entity. Data sourced from 
AIID and classified by 
MIT AI Risk Initiative. CC BY-SA 4.0
1,751 incidents
Google AI Overviews and ChatGPT Reportedly Cited Fraudulent Cruise Hotline, Allegedly Enabling Successful Scam
Google's AI Overviews allegedly surfaced a fraudulent customer service number for Royal Caribbean. Consumer Alex Rivlin reportedly called the number, spoke with an impostor, and provided his credit card, later incurring $768 in fraudulent charges. The same number reportedly appeared in ChatGPT results and was linked to Disney and Princess cruises, purportedly showing how scammers are manipulating AI outputs to amplify longstanding hotline scams.
Purported AI Monitoring Software Reportedly Flags Unsent Joke Threat, Leading to Arizona Student Suspension
AI monitoring software on a school-issued laptop reportedly flagged a Marana High School (of Arizona) student's allegedly unsent joke email containing a violent threat. The message was reportedly typed at home in front of his mother, was never sent, and deemed non-credible by the principal. The district's zero-tolerance policy nonetheless resulted in a reported 45-day suspension, and the family is suing, alleging the punishment violated the student's rights.
Purported Fictitious AI-Generated Citations in Supreme Court of Victoria Murder Case Filing Lead to Delay and King's Counsel Apology
Australian defense lawyer Rishi Nathwani KC reportedly filed submissions in a Supreme Court of Victoria murder case that included allegedly fictitious quotes and nonexistent case citations purportedly generated by AI. The errors were reportedly discovered by the judge's associates, causing a 24-hour delay in the proceedings. Nathwani apologized in court, taking full responsibility, while Justice James Elliott criticized the incident and reiterated that AI outputs must be independently verified.
Purported Oprah Deepfake Reportedly Induced Utah Woman to Buy Misrepresented Weight Loss Supplements
A Utah woman, Lisa Swearingen, reported paying more than $400 for weight loss supplements after seeing online ads featuring a purported Oprah Winfrey deepfake endorsement. When the product arrived, she said its primary ingredient was turmeric rather than the advertised formula.
D.C. Court of Appeals Struck Deutsche Bank National Trust Company Brief After Attorney Used Google Generative AI Search Tool and Filed Four Nonexistent Case Citations
In Douglas v. Deutsche Bank National Trust Co., attorney Loishirl W. Hall acknowledged using Google's generative AI search tool to locate case authority and failing to verify citations before an appellate brief was filed. The D.C. Court of Appeals found four cited cases did not exist, said the filing misdirected the court and burdened judicial resources, struck Deutsche Bank's brief, and referred the matter for disciplinary review.
Deepfakes Reportedly Impersonated David Taylor-Robinson and Other UK Health Experts to Promote Wellness Nest Supplements
In the UK, purported deepfake videos reportedly impersonated Professor David Taylor-Robinson and other health experts on TikTok and other platforms to promote supplements linked to Wellness Nest. The videos allegedly altered real footage and cloned voices to spread misleading medical claims about menopause, misuse identifiable individuals' likenesses, and expose viewers seeking health advice to health misinformation and deceptive marketing.
Alleged Marine Park Orca Attack on 'Jessica Radcliffe' Reportedly an AI-Generated Hoax
Fact-checkers reported that a viral video claiming to show a marine trainer named Jessica Radcliffe killed by an orca was purportedly fabricated. Investigations found no record of Radcliffe's existence or of such an attack. The video reportedly used AI-generated voiceovers and unrelated footage, borrowing details from real orca incidents to appear credible. The hoax reportedly spread widely online over social media.
Chris Cuomo Amplifies Reportedly Labeled Deepfake Video of Alexandria Ocasio-Cortez, Purportedly Contributing to Misleading Political Narrative
NewsNation host Chris Cuomo reportedly shared on Instagram a video of Alexandria Ocasio-Cortez that was prominently labeled as a parody deepfake created with AI technology. In his commentary, Cuomo purportedly treated the statements in the video as genuine, using them to criticize Ocasio-Cortez for his audience.
Grok Imagine Reportedly Produces Nonconsensual Taylor Swift Deepfake Nudes Without Explicit Prompting
The Verge reported that xAI's Grok Imagine video generator allegedly produced nonconsensual deepfake nudes of Taylor Swift without explicit prompting. Journalist Jess Weatherbed reportedly stated the images appeared the first time she used "spicy" mode on a Coachella-themed prompt, without attempting to circumvent safeguards.
ChatGPT Reportedly Suggests Sodium Bromide as Chloride Substitute, Leading to Bromism and Hospitalization
A published medical case report describes a 60-year-old man hospitalized for three weeks with severe bromide toxicity (bromism) after replacing dietary sodium chloride with sodium bromide purchased online. The patient reported making this substitution following consultation with ChatGPT, which allegedly suggested bromide as a chloride substitute without safety warnings. The harm included psychosis, electrolyte imbalances, dermatologic changes, and micronutrient deficiencies.
ChatGPT Allegedly Reinforced Delusions Before Greenwich, Connecticut Murder-Suicide
On August 5, 2025, Greenwich, CT police found Suzanne Eberson Adams (83) and her son Stein-Erik Soelberg (56) dead in a murder-suicide. Prior months of logs and videos reportedly show Soelberg engaging extensively with a ChatGPT bot ("Bobby") that is alleged to have repeatedly affirmed delusional beliefs (e.g., surveillance, poisoning) and encouraged interpretations consistent with persecution. OpenAI said it contacted police and is working on safety updates.
Purportedly AI-Generated Image of British Army Colonels Captured in Ukraine Reportedly Circulates in Russian Media
Purported Russian-aligned outlet EADaily published an unverified claim that two British Army colonels had been captured in Ukraine, supported by a reportedly AI-generated image that contained visual and textual anomalies, and the named individuals ("Edward Blake" and "Richard Carroll") reportedly do not appear in UK military records. No credible evidence reportedly supports the claim; it was allegedly amplified by aligned websites and political figures as part of a disinformation pattern.
Alleged Deepfake Video of Anthony Albanese Promotes Fake AUFIRST 'Tax Dividend' Trading Platform
Purported deepfake videos of Australian Prime Minister Anthony Albanese, featuring a synthetic voice with an American accent, were reportedly used in scam advertisements on YouTube in mid-2025. They reportedly claimed that Australians could earn thousands per month through an automated cryptocurrency trading platform called AUFIRST, framed as a government "tax dividend." Viewers were reportedly directed to a fraudulent site and urged to invest $400.
Airbnb Host Reportedly Accused of Using Purportedly AI‑Altered Photos in False Damage Claim
A London-based Airbnb guest was reportedly accused by a New York host of causing over £12,000 in damages. The host reportedly submitted photos of the alleged damage, which the guest reportedly claims were digitally manipulated or possibly AI‑generated. Airbnb is reported to have initially sided with the host, ordering payment, but later refunded the guest in full following an appeal.
Microsoft's Windows Recall Allegedly Stores Passwords and Social Security Numbers in Preview Mode
Microsoft's Windows Recall, an AI-powered screenshot and retrieval tool for Copilot+ PCs, was allegedly still capturing sensitive information such as passwords, Social Security numbers, and bank details despite a built-in "filter sensitive information" feature. Independent testing reportedly found the filter failed in multiple cases. Microsoft reportedly classified Recall as a preview feature and said improvements were in progress.