Video showing moon interpreted as yellow stop light
An intriguing video clip showing the moon in its full glory was recently misconstrued as a yellow traffic light by an AI model, highlighting...
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 & citation Failure Files™Discrimination through optimization: How Facebook's ad delivery can lead to skewed outcomes
An analysis of Facebook's ad delivery system reveals potential for discrimination by optimizing content to specific user demographics. Under...
Read moreThe English test that ruined thousands of lives
A popular AI-based English language test has been under scrutiny for producing inaccurate results, potentially ruining thousands of lives. T...
Read moreMajor fire breaks out at Ocado warehouse after three robots collide
In an unexpected turn of events, a fire erupted at the Ocado warehouse after three robots inadvertently collided. This incident underscores...
Read moreFacebook’s race-blind practices around hate speech came at the expense of Black users, new documents show
Recent documents unveil that Facebook's policies regarding hate speech were race-blind, leading to a disproportionate impact on the platform...
Read moreFacebook Tried to Make Its Platform a Healthier Place. It Got Angrier Instead.
An in-depth analysis of Facebook's efforts to foster a healthier platform using AI, focusing on the unintended consequences that led to incr...
Read moreOhio Sues Meta Alleging Facebook Parent Misled Public About Its Products’ Effect on Children
Ohio has filed a lawsuit against Meta Platforms Inc., the parent company of Facebook, claiming that it misled the public about the effect of...
Read moreWhat a machine learning tool that turns Obama white can (and can’t) tell us about AI bias
An exploration of a machine learning tool that manipulates images to portray former President Obama as white, revealing both its limitations...
Read moreThe Best Algorithms Still Struggle to Recognize Black Faces
An exploration of the ongoing challenge faced by current algorithms in accurately recognizing Black faces, emphasizing the need for responsi...
Read moreThis girls-only app uses AI to screen a user’s gender — what could go wrong?
An app designed for girls-only uses AI to screen users' gender, raising concerns about responsible AI implementation, data privacy, and pote...
Read moreWhy Stanford Researchers Tried to Create a ‘Gaydar’ Machine
In an attempt to advance Artificial Intelligence (AI) research, Stanford researchers embarked on a project aimed at developing an AI system...
Read moreWhy AI Isn’t Providing Better Product Recommendations
Exploring why AI's performance in delivering accurate and beneficial product recommendations falls short, emphasizing the importance of resp...
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.