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
741 incidents
NullBulge's AI-Powered Malware Allegedly Compromises Disney Employee and Internal Data
A Disney employee, Matthew Van Andel, reportedly downloaded AI-powered malware allegedly developed by the cybercriminal group NullBulge, resulting in a major cybersecurity breach. Hackers purportedly accessed Disney's Slack system, exposing 44 million internal messages, employee and customer data, and financial records. NullBulge also reportedly leaked Van Andel’s personal financial information, leading to identity theft and his eventual termination.
Department for Work and Pensions (DWP) AI Systems Allegedly Discriminate Against Single Mothers
Researchers have argued that the Department for Work and Pensions' Universal Credit system disproportionately impacts single mothers. Automated processes in the system, designed to determine eligibility and detect fraud, are reported to have introduced biases, leading to financial instability and hardship. The algorithms allegedly miscalculate earnings and delay childcare reimbursements, in turn exacerbating income volatility and debt among single mothers.
Baidu Robotaxi Allegedly Involved in Collision with Pedestrian in Wuhan
On July 7, 2024, a Baidu Robotaxi reportedly collided with a pedestrian at a traffic intersection in Wuhan. The incident occurred as the autonomous vehicle started moving on a green light while the pedestrian was allegedly crossing against a red light. The pedestrian sustained minor injuries, and Baidu was reported to have been cooperating with local authorities for further investigation.
Figma Disables AI Feature After Accusations of Copying Apple’s Weather App
Figma has temporarily disabled its AI design feature, "Make Design," after accusations of copying Apple’s Weather app. Andy Allen of NotBoring Software highlighted the issue, prompting Figma CEO Dylan Field to deny claims of training the AI on specific app designs. However, Field acknowledged flaws in the QA process and promised to suspend the feature until it meets quality standards. The incident has implications for designers.
OpenAI's ChatGPT Mac App Stored User Data in Unencrypted Text Files
OpenAI's ChatGPT macOS app stored user conversations in plain text. If accessed by a malicious actor, these conversations could have been easily read. The critical security flaw was demonstrated by a third party and ultimately resolved after OpenAI released an update to encrypt the stored data.
Students at Australian Schools Reportedly Used AI Nudification Tools to Create Nonconsensual Sexual Images of Classmates and Teachers
Throughout 2024, schools in Australia reportedly dealt with a significant rise and proliferation of nonconsensual deepfake pornography of students. Often, male students are reported to use "nudify" apps such as Undress AI with images of their classmates and teachers. Many of the sites have reportedly remained legal and allegedly accessible to minors, who in turn are purportedly using the sites to generate pornography of their peers.
AI Work Assistants Require More Effort Than Expected, CIOs Say
AI work assistants, such as Copilot for Microsoft 365 and Gemini for Google Workspace, are proving to be more labor-intensive than anticipated for enterprises. CIOs report that these AI tools struggle with outdated or inaccurate data, leading to incorrect outputs. Companies are finding they must invest heavily in data management to ensure reliability. This added effort has led to delays in deployment and frustration, as businesses work to maximize the potential of these expensive AI tools.
Department for Work and Pensions (DWP) Algorithm Wrongly Flags 200,000 for Housing Benefit Fraud
A Department for Work and Pensions (DWP) algorithm wrongly flagged over 200,000 UK housing benefit claims as high risk, resulting in unnecessary investigations. Two-thirds of these flagged claims were legitimate, causing wasted public funds and stress for claimants. Despite initial success in a pilot, the algorithm's real-world performance fell short. This incident highlights the risks of overreliance on automated systems in welfare administration.
Unrestricted AI Avatar Tool Accidentally Released by TikTok Permits Recitation of Hitler Quotes and Other Harmful Speech
TikTok accidentally released an internal version of its AI digital avatar tool without safeguards, allowing users to generate videos where avatars could recite harmful content, including quotes from Hitler. The tool, meant for advertisers, was accessible to personal accounts and lacked the watermark indicating AI-generated content. TikTok has since removed the tool and acknowledged the problem.
Instagram's Algorithm Reportedly Recommended Sexual Content to Teenagers' Accounts
Tests by The Wall Street Journal and a researcher reportedly revealed that Instagram's AI-driven Reels algorithm recommends sexually suggestive content to accounts listed as 13 years old. Despite Meta's commitment to restrict such content for minors, explicit videos were served within minutes of account creation, according to the findings.
Erroneous Declined Transaction Notification by PayPal AI Assistant
On July 13th, 2024, a user reported an incident involving PayPal's generative AI chatbot. The chatbot allegedly incorrectly informed the user of a declined transaction that never occurred, causing confusion and prompting a call to customer service for clarification. This false alert suggests a flaw in the AI system's reliability. The incident created unnecessary labor for both the user and PayPal's human support, demonstrating the potential harm of deploying generative AI without thorough testing and error handling mechanisms.
Leading AI Models Reportedly Found to Mimic Russian Disinformation in 33% of Cases and to Cite Fake Moscow News Sites
An audit by NewsGuard revealed that leading chatbots, including ChatGPT-4, You.com’s Smart Assistant, and others, repeated Russian disinformation narratives in one-third of their responses. These narratives originated from a network of fake news sites created by John Mark Dougan (Incident 701). The audit tested 570 prompts across 10 AI chatbots, showing that AI remains a tool for spreading disinformation despite efforts to prevent misuse.
Tesla Reportedly in Autopilot Mode Hits Parked Police Vehicle in Fullerton, California
A Tesla reportedly in self-driving mode crashed into a parked patrol vehicle in Fullerton, California while the officer was responding to a fatal DUI crash. The officer narrowly escaped injury. The driver reports having been distracted by a cellphone and having relied on the Tesla’s AI. (The earlier crash involved a suspected DUI driver who killed a motorcyclist stopped at a red light.)
Independent News Sites Flagged as Spam by Facebook's AI Moderation System
Facebook's AI moderation system mistakenly flagged and removed legitimate news articles from independent local publishers as spam. This affected publishers worldwide, disrupting content distribution and impacting traffic. Despite attempts to appeal, many publishers received no response.
BNN Breaking's AI-Driven Errors Reportedly Damage Reputations and Spread Misinformation
BNN Breaking, an AI-driven news site, published a false story about Irish DJ Dave Fanning, damaging his reputation. The site used AI to generate error-filled content, leading to numerous complaints and a defamation lawsuit against BNN and Microsoft. The site has since gone dormant.