AI-Powered Fraud: Impersonation of UK Executive's Boss Reveals Vulnerabilities
Mar 1, 2019
AI voice technology was reportedly manipulated by fraudsters to mimic the boss of a UK firm's CEO, attempting a €220,000 wire transfer over...
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
Data source & citationOpenAI's GPT-2 Memorizes PII in Training Data: A Case for Safe and Secure AI
Feb 14, 2019
The incident involving OpenAI's GPT-2 highlights the importance of trustworthy AI. The model memorized and could recite verbatim instances o...
Read moreGraphic Content on YouTube: Failure of AI-Based Moderation Systems in Regulating Suicide and Self-harm Videos
Feb 4, 2019
Terms-of-service-violating videos related to suicide and self-harm reportedly bypassed YouTube’s AI-based content moderation systems, potent...
Read moreClimate Misinformation: Alleged Impact of YouTube's Recommendation System on Viewers
Feb 1, 2019
An advocacy group claims that YouTube's recommendation algorithm, prioritizing views and watched time, may have inadvertently guided users t...
Read moreApple Maps Incident: Unsafe Navigation Raises Concerns for Responsible AI
Feb 1, 2019
Near Los Angeles, Apple Maps allegedly misguided a ski-trip couple onto an unconventional route, leading them to get lost and stranded on an...
Read moreWrongful Arrest: The Importance of Responsible AI in Facial Recognition Technology - New Jersey Case
Jan 30, 2019
The Woodbridge Police Department inadvertently arrested an innocent individual, highlighting the need for safe and secure AI governance in f...
Read moreSEC Charges Two Investment Firms for Misrepresenting Artificial Intelligence Capabilities: Harm Prevention in AI Governance
Jan 1, 2019
In an instance of AI washing, the SEC has accused Delphia and Global Predictions for misstating their utilization of AI in their investment...
Read moreIncident: Warehouse Robot Causes Bear Spray Accident, Injuring 24 Amazon Workers - Emphasizing Harm Prevention in AI Governance
Dec 5, 2018
Twenty-four workers at an Amazon warehouse in New Jersey were injured after a robot accidentally punctured a can of bear repellent spray. Th...
Read moreMisclassification of Inoffensive Images by Tumblr's AI-Powered Adult Content Filter: A Case for Responsible AI Governance
Dec 3, 2018
Tumblr’s AI-powered adult content filter mistakenly flagged inoffensive images as explicit following its ban on all adult content. This inci...
Read moreDubious Facial Recognition Pilot in Bahia: Disparate Impact on Black and Low-Income Populations
Dec 1, 2018
A facial recognition pilot project implemented by the local government of Bahia, reportedly with a minimal hit rate, appears to have disprop...
Read moreSleepy Driver: A Tesla Autopilot Incident Highlighting Need for Responsible AI
Dec 1, 2018
A Tesla Model S incident occurred when an intoxicated, asleep driver failed to monitor the vehicle on autopilot. Police intervention was req...
Read moreUnreliable and Imbalanced Schufa Credit Scoring in Germany: A Case for Responsible AI
Nov 28, 2018
This AI incident involving Schufa credit scoring in Germany highlights the importance of trustworthy and fair AI. The scores reportedly priv...
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.