Twitter's AI Moderation Tool Misidentified Rockets as Pornography
Jan 3, 2023
Twitter's automated content moderation tool inadvertently misclassified images of rocket launches as explicit content, leading to incorrect...
Read moreValues Statement: We believe AI should cause no harm, but enhance the quality of human life, by proactively adopting our AI Governance framework.
Evidence-based Transparent For governance
AI Incidents
TAIM catalog Map Data source & citationAI-Generated-Text-Detection Tools Reported for High Error Rates
Jan 3, 2023
Recent reports reveal high error rates in models designed to detect text generation AI, such as AI Text Classifier and GPTZero. These tools...
Read moreShotSpotter Failed to Alert Authorities of Mass Shooting in North Carolina
Jan 1, 2023
The ShotSpotter system failed to detect gunshots during a drive-by shooting in Durham, North Carolina on New Year's Day, resulting in five i...
Read moreAI-Powered VTuber and Virtual Streamer Made Toxic Remarks on Twitch
Dec 28, 2022
An LLM-powered VTuber and streamer on Twitch made provocative statements, including Holocaust denial, the assertion that women's rights do n...
Read moreTesla Driver Put Car on Autopilot Before Falling Asleep in Germany
Dec 28, 2022
A Tesla driver on the Autobahn near Bamberg, Germany, activated his vehicle's Autopilot mode before falling asleep. Despite efforts by the p...
Read moreChatGPT Abused to Develop Malicious Softwares
Dec 21, 2022
Cyber criminals exploited OpenAI's ChatGPT, using it to create malware, ransomware, and other harmful software. This incident underscores th...
Read moreSouthwest Airlines Crew Scheduling Solver Degenerates Flight Network
Dec 21, 2022
Southwest Airlines encountered an unexpected crisis when its crew scheduling software repeatedly malfunctioned in response to weather-induce...
Read moreFootballer's "X-Rated" Comment Created by Instagram's Mistranslation
Dec 19, 2022
A footballer's comment, originally in Spanish, was mistranslated by Instagram to English, resulting in a message that appeared 'racy' and 'X...
Read moreLawyers Denied Entry to Performance Venue by Facial Recognition
Dec 19, 2022
Lawyers were prevented from entering Madison Square Garden after a facial recognition system incorrectly identified them as being employed b...
Read moreStartup Misled Research Participants about GPT-3 Use in Mental Healthcare Support
Dec 1, 2022
OpenAI's powerful language model, GPT-3, was used by a mental health startup for peer-to-peer support without proper ethical review. The int...
Read moreAI Video-Making Tool Abused to Deploy Pro-China News on Social Media
Dec 1, 2022
An incident involving Synthesia's AI-powered video tool, abused by Spamouflage to spread pro-China propaganda on social media, underscores t...
Read moreFacebook's Job Ad Algorithm Allegedly Biased against Older and Female Workers
Dec 1, 2022
The HISPI Project Cerebellum AI incident database highlights an allegation that Facebook's algorithm selectively showed job advertisements d...
Read moreData source
Incident data is from the AI Incident Database (AIID).
When citing the database as a whole, please use:
McGregor, S. (2021) Preventing Repeated Real World AI Failures by Cataloging Incidents: The AI Incident Database. In Proceedings of the Thirty-Third Annual Conference on Innovative Applications of Artificial Intelligence (IAAI-21). Virtual Conference.
Pre-print on arXiv · Database snapshots & citation guide
We use weekly snapshots of the AIID for stable reference. For the official suggested citation of a specific incident, use the “Cite this incident” link on each incident page.