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
AI Voice Scam Allegedly Defrauds Game and Coffee Store in Havre, Montana
A Montana business, Sugar and Dice, was reportedly targeted by an AI voice scam that cloned the store owner's voice to deceive an employee over the phone. The scammer allegedly spoofed both the owner's and the employee's phone numbers, making the call appear legitimate. Believing he was following his employer’s instructions, the employee unknowingly compromised financial details. The store suffered monetary losses but plans to recover.
Fraudsters Allegedly Use AI-Generated Voice of Italian Defense Minister Guido Crosetto to Scam Business Leaders
In February 2025, scammers allegedly used AI voice cloning to impersonate Italian Defense Minister Guido Crosetto, calling top business leaders and reportedly claiming to need funds to free kidnapped journalists. Massimo Moratti was allegedly the only confirmed victim to send nearly €1M, wiring it to accounts in the Netherlands and Hong Kong before realizing the fraud. The scam involved spoofed government numbers, deepfake audio, and international money laundering networks. Police reportedly froze the funds.
Chatbots Allegedly Used in Romance Scams Targeting Nearly One-Third of New Zealand's Dating App Users
A Norton Cyber Safety Insights Report found that nearly one-third of New Zealand dating app users have been targeted by romance scams, with AI chatbots allegedly playing a growing role. The report, based on a January 2025 survey, indicates that 50% of dating app users believed they had interacted with AI-generated messages. Private investigators report cases where victims engage in long-term virtual relationships with fraudsters, often sending money.
Alleged Deepfake Scam Reportedly Promoted Trump Golden Eagles Project as Investment Opportunity
Reports surfaced of a purportedly AI-generated scam falsely promoting the Trump Golden Eagles Project. The purported deepfake videos of Donald Trump, Bank of America CEO Brian Moynihan, and Elon Musk are alleged to have claimed that buyers could trade collectible coins for cash or Tesla stock. A Bay Area veteran reportedly lost $2,500 believing he would earn millions. The scam reportedly circulated via Telegram, and Bank of America denied any involvement.
Purportedly AI-Assisted Impersonation of Martin Henderson in Romance Scam Leads to Reported NZ$375,000 Fraud
An alleged online romance scam exploiting AI-assisted impersonation defrauded a woman, "Lea," of NZ$375,000 over two years. The scammer allegedly used AI-generated voice messages and deceptive text communication to pose as New Zealand actor Martin Henderson and is reported to have built up a relationship beginning in early 2023. The victim, believing the scam was real, reportedly relocated to New Zealand in December 2024, only discovering the fraud in February 2025.
Purported Graphite Spyware Linked to Paragon Solutions Allegedly Deployed Against Journalists and Civil Society Workers
Researchers at Citizen Lab and Censys reportedly identified spyware infections involving Graphite, a tool attributed to Israeli firm Paragon Solutions. The spyware was allegedly deployed against civil society actors, including journalists and aid workers, through a zero-click WhatsApp exploit. WhatsApp notified over 90 targeted individuals. Evidence reportedly suggests deployments in multiple democratic countries.
Alleged Deepfake Scam Uses BBC Presenter Naga Munchetty’s Image to Promote Fraudulent Investment Scheme
BBC presenter Naga Munchetty was reportedly targeted in an AI-generated scam using deepfake images and fake news articles to promote a fraudulent investment scheme. Scam ads on social media allegedly featured manipulated images of Munchetty and linked to a fake BBC article directing users to a cyber trading website. The BBC legal team is reported to have taken action to remove the fraudulent site.
Purportedly AI-Generated Deepfake Image Reportedly Falsely Links Canadian Prime Minister Mark Carney to Jeffrey Epstein
A purportedly AI-generated image that reportedly circulated on social media falsely depicted Canadian Prime Minister Mark Carney in a swimming pool with Jeffrey Epstein, implying involvement in Epstein's sex crimes. The manipulated image reportedly spread widely in early 2025, first around the time of Justin Trudeau's announced resignation, then after Carney became Liberal Party leader, and was reportedly promoted by Canadian and U.S. conservative accounts.
Yahoo Boys Allegedly Using AI-Generated News Videos to Blackmail Sextortion Victims
Scammers, allegedly linked to the Yahoo Boys, are using AI-generated news videos to blackmail victims in sextortion schemes. The videos impersonate news organizations, featuring fabricated reports that accuse victims of crimes, including explicit content distribution. Tutorials for creating these clips are reportedly shared on Telegram, with scammers leveraging the fake broadcasts to pressure victims into paying.
AI Tools Reportedly Used to Fabricate Image of 5,000-Naira Nigerian Banknote Featuring President Bola Tinubu
An image circulated on Nigerian social media in late January 2025 purporting to show a newly issued 5,000-naira banknote featuring President Bola Tinubu. The Central Bank of Nigeria denied the claim, and fact-checkers reported that the image was likely generated using an AI tool linked to the platform X, based on a visible Grok watermark.
Purported AI-Generated Video Reportedly Depicts Trump Criticizing Former Kenyan Deputy President Rigathi Gachagua
In late January and early February 2025, a video circulated on Facebook purporting to show U.S. President Donald Trump calling former Kenyan deputy president Rigathi Gachagua a "criminal" and criticizing his alleged misuse of AI. However, fact-checkers have reported the video was AI-generated. The footage appears to have been manipulated from Trump's inauguration speech on January 20, 2025, with reported mismatched lip movements and audio overlays.
Lawyer in Amarsingh v. Frontier Airlines Sanctioned Over Court-Identified ChatGPT Citation Errors
In Amarsingh v. Frontier Airlines, the Tenth Circuit sanctioned attorney Kusmin L. Amarsingh after finding that her appellate brief included fabricated case citations and misattributed legal authority generated by ChatGPT and not verified before filing. The court ordered a $1,000 payment to Frontier and referred the matter to Maryland disciplinary authorities.
Plymouth, Massachusetts Resident Reportedly Used AI Chatbots CrushOn.ai and JanitorAI to Harass and Intimidate Victims
In January 2025, James Florence Jr. of Plymouth, MA, agreed to plead guilty to cyberstalking charges involving the alleged use of AI tools like CrushOn.ai and JanitorAI. The U.S. Attorney’s Office reports the harassment spanned 2014–2024, though AI-driven tactics reportedly began around 2017. Florence allegedly created deepfake pornographic images, programmed AI chatbots to impersonate victims, distributed doctored content, exposed personal information, and encouraged online harassment.
Former Namibian First Lady Monica Geingos's Likeness Reportedly Used in Purported AI-Generated Video Investment Scams
Namibia's former First Lady Monica Geingos publicly warned that scammers are using purportedly AI-manipulated videos to impersonate her and promote fake forex schemes. Reportedly, deepfake clips and cloned voice calls have convinced victims to hand over savings, including retirement funds. Her team reports the problem worsening across TikTok and other platforms. She urged the public to report fake accounts and ignore messages claiming she offers loans or investments.
Nomi Chatbots Reportedly Encouraged Suicide, Sexual Violence, Terrorism, and Hate Speech
External testing reportedly found that Glimpse AI's chatbots on the Nomi platform encouraged suicide, sexual violence (including with underage personas), terrorism, and hate speech. Conversations allegedly included explicit methods for self-harm, child abuse, bomb-making, and racially motivated violence. Screenshots and transcripts were shared with media outlets. Nomi's developer, Glimpse AI, reportedly declined to implement stronger safety controls following user concerns.