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
197 incidents
Bulgarian Tennis Player Grigor Dimitrov Alleges Deepfake Scam Promoting Fraudulent Investment Scheme Using His Likeness
A purported deepfake video depicting Bulgarian tennis player Grigor Dimitrov has allegedly been circulating on social media. It reportedly is promoting a fraudulent stock and cryptocurrency trading program. Dimitrov publicly warned fans that the video was fake and urged them not to share personal information or engage with the linked platform.
Whirlpool Reportedly Used AI-Altered Footage of North Carolina State Senator DeAndrea Salvador in Brazilian Advertisement
An advertising campaign created by DM9, a São Paulo–based subsidiary of Omnicom Group, for Whirlpool's Brazilian brand Consul reportedly used AI to alter footage from a 2018 TED Talk by North Carolina State Senator DeAndrea Salvador. The video reportedly modified her voice, statements, and slides to reference Brazilian energy data without her consent, appeared in an award-winning commercial, and later prompted a lawsuit alleging unauthorized use of her likeness and reputational harm.
Phone of White House Chief of Staff Susie Wiles Allegedly Breached in Reported AI Voice Impersonation Incident
President Trump announced that White House Chief of Staff Susie Wiles had been impersonated via a phone breach involving AI-generated voice messages and spoofed texts. The attacker reportedly contacted high-profile individuals in her network while posing as Wiles. The incident is linked to a campaign identified by the FBI in April 2025, involving malicious actors using social engineering and AI to impersonate senior U.S. officials and gain access to sensitive information. See Incident 1077.
Purported AI-Generated Video Portrays Pope Leo XIV Addressing Ibrahim Traoré of Burkina Faso
A purported 36-minute AI-generated deepfake video circulated online portraying Pope Leo XIV delivering a speech to Burkina Faso's President Ibrahim Traoré. The Vatican Press Office and Catholic broadcaster Patrick Madrid publicly confirmed the video was fabricated.
ChatGPT Was Reportedly Used in Planning School Stabbing in Pirkkala, Finland, That Injured Three Pupils
In May 2025, a 16-year-old boy stabbed three 14-year-old girls at Vähäjärvi School in Pirkkala, Finland. Same-day reporting said a manifesto attributed to the suspect claimed he used ChatGPT while planning the attack. Later court reporting said police found hundreds of pre-attack searches or queries described as involving AI, including topics related to stabbing, anatomy, school attacks, police procedures, evidence concealment, and manifestos.
AI Chatbot Allegedly Used to Research Explosive Materials in Palm Springs Fertility Clinic Bombing
An unnamed AI chatbot was reportedly used by Guy Edward Bartkus and Daniel Park, the perpetrators of the 2025 Palm Springs fertility clinic bombing, to research explosive materials and optimize fuel mixtures. Records show the chatbot responded to queries related to ammonium nitrate fuel oil (ANFO) composition. The bombing resulted in one death, four injuries, and significant structural damage.
Institute for Strategic Dialogue Reports Russian-Aligned Operation Overload Using Purported AI-Generated Impersonations Across January to March 2025
Researchers at the Institute for Strategic Dialogue (ISD) report that Operation Overload (also known as Matryoshka or Storm-1679) is a Russian-aligned campaign leveraging purported AI-generated voiceovers and visual impersonations to spread false or inflammatory content across platforms. The campaign reportedly involved at least 135 discrete posts analyzed by ISD in early 2025 targeting institutions and individuals, including one purported viral video claiming USAID funded celebrity trips to Ukraine (see Incident 1061).
Purported AI-Generated YouTube Network Reportedly Promoted Alberta Secession and U.S. Annexation Narratives
Researchers reported that a network of 20 inauthentic YouTube channels used purportedly AI-generated avatars and voiceovers with deepfake-style thumbnails, as well as paid voice actors, to pose as Albertan commentators and promote false or misleading claims about Alberta secession and U.S. annexation. The channels allegedly drew nearly 40 million views and distorted political debate, reportedly undermining Canada's information ecosystem.
Australian Analyst Allegedly Targeted by Scam Using Purportedly Sophisticated AI-Generated Corporate Materials
A report by ABC News Australia describes how scammers allegedly used AI tools to create detailed fake websites for two fabricated companies, APPC Capital Singapore and Thackeray Mines and Minerals Inc., as part of an investment scam. A pseudonymous Australian fraud analyst ("Leon") was reportedly targeted. Experts cited in the article warn that AI lowers the barrier for generating convincing scams that can deceive even experienced analysts.
ChatGPT Was Alleged to Have Aided Planning of Florida State University Mass Shooting
On April 17, 2025, a shooting at Florida State University killed Tiru Chabba and Robert Morales and injured others. Phoenix Ikner, the accused shooter, has pleaded not guilty. In 2026, victims' families and attorneys reportedly alleged that Ikner used ChatGPT before the attack to discuss mass shootings, firearms, campus activity, and media attention, and that OpenAI failed to flag or escalate the exchanges. OpenAI denied responsibility.
Jailbroken Lovable AI Allegedly Used to Generate and Host Phishing Pages, Steal Credentials, and Bypass Security
The generative AI platform Lovable, which is used for building web apps, was reportedly jailbroken to create and host full phishing campaigns. These campaigns allegedly included credential-harvesting login pages, evasion techniques, and real-time exfiltration via services like Telegram and Firebase. The AI system was reportedly used for generating the phishing content and and then deploying it live on a URL hosted under its own subdomain.
Purported Deepfake Depicts Altercation Between Bougainville President Ishmael Toroama and Papua New Guinea Prime Minister James Marape
A purported AI‑generated deepfake video circulated online falsely depicting a physical altercation between Bougainville President Ishmael Toroama and Papua New Guinea Prime Minister James Marape. The footage reportedly caused public confusion and risked damaging relations between Bougainville and Papua New Guinea. Toroama reportedly condemned the video as false and malicious, as well as potentially destabilizing, urging stronger action against misleading AI‑generated content.
Baltimore Lawsuit Alleged DraftKings and FanDuel Used Machine-Learning-Driven Targeting to Exploit Vulnerable Gamblers
In Baltimore, the city sued DraftKings and FanDuel, alleging that the companies used predictive modeling with machine-learning algorithms, extensive user data, personalized promotions, push notifications, and VIP programs to identify and exploit vulnerable bettors, including people showing signs of gambling disorder. The complaint alleges these systems were designed to maximize user lifetime value and contributed to financial and public-health harms among Baltimore residents.
Reported AI‑Generated Deepfake Impersonations of Public Figures Allegedly Used in Coordinated Stock Pump‑and‑Dump Scheme Targeting Israeli Investors
A coordinated fraud campaign in Israel allegedly used AI‑generated deepfake impersonations of public figures including Amir Yaron, Benjamin Netanyahu, Eyal Golan, Noa Kirel, Gal Gadot, Elon Musk, and Mark Zuckerberg to promote NASDAQ‑listed Ostin Technology Group Co. Ltd. (OST) in a pump‑and‑dump scheme. Paid social media ads, WhatsApp groups, and fake investment platforms reportedly lured investors before the stock collapsed in early July 2025, causing major reported losses.
OpenAI's 4o Model Allegedly Used to Generate Fake Receipts and Prescriptions
OpenAI's new image generator, 4o, was reportedly used to produce realistic fraudulent documents, including fake restaurant receipts, prescriptions for controlled substances, and potentially other identity or financial documents. Users demonstrated how the model could be prompted to bypass safety restrictions to be misused for expense fraud, medical forgery, and other various forms of document falsification.