Values 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

This Software Company's Transformation Towards Trustworthy AI: Addressing Recent Controversies

Exploring changes made by the software company in response to recent controversies, emphasizing their commitment to safe and secure AI. This...

Analyzing the COMPAS Recidivism Algorithm: A Step Towards Trustworthy AI

Delve into our analysis of the controversial COMPAS recidivism algorithm, shedding light on its impact and highlighting the importance of sa...

Debiasing Word Embeddings: Advancing Responsible AI for Gender Equality

Explore the importance of debiasing word embeddings in AI, a key aspect of trustworthy AI. This approach helps prevent harmful gender stereo...

Google's Anti-Internet Troll AI Platform Vulnerable: A Lesson in Safe and Secure AI

Security researchers expose a potential flaw in Google's anti-internet troll AI platform, highlighting the importance of trustworthy AI and...

Google's AI Incident: Inappropriate Opinions on Homosexuality - The Importance of Safe and Responsible AI

This incident underscores the need for trustworthy AI, as Google's AI was found to have inappropriate opinions about homosexuality. By joini...

Amazon's Algorithm Adjustments for Harm Prevention: A Case Study on GLBT Books

In this article, we discuss Amazon's move to censor search results and rankings related to GLBT books. This AI incident maps to the Govern f...

Racially Biased AI Incident: Google Apologizes for Auto-Tagging Error in Photo App

Google acknowledges an incident involving its photo app, where images of Black people were auto-tagged incorrectly. This AI incident maps to...

Exploring Google's Emotional Email-Replying AI: A Case Study on AI Governance

Delve into an analysis of Google's emotional email-replying AI incident, a valuable insight into AI governance and the need for trustworthy...

Exploring and Addressing Unseen Gender Bias in Google Image Search - A Step Towards Responsible AI

Delve into the lesser-known gender bias within Google Image Search, a crucial aspect of AI governance. This incident maps to the Govern func...

Investigation of Amazon Warehouse Bear Spray Accident Highlights Importance of Safe and Secure AI in Work Environments - OSHA

The Occupational Safety and Health Administration (OSHA) is currently investigating an incident at an Amazon warehouse involving bear spray,...

Exploring Potential Risks in Google's Ad Targeting System: A Case for Responsible AI

This incident sheds light on potential issues within Google's ad-targeting system, emphasizing the need for safe and secure AI. It maps to t...

Autopilot Malfunction: Tesla Crashes into Police Car While Driver Watches Movie - An Incident Illustrating the Need for Responsible AI

In this alarming incident, a Tesla vehicle equipped with Autopilot collided with a police car. The driver admitted to watching a movie at th...

Data 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.