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

Uncovering Bias in AI Chest X-Ray Classification: A Call for Responsible AI

Recent research reveals racial, gender, and socioeconomic bias in chest X-ray classifiers. This AI incident maps to the Govern function in H...

The Lekki Massacre: Investigating the Misinformation Surrounding the October 20 Incident on Facebook - A Call for Responsible AI

Explore this critical incident that led to Facebook's labeling of content related to the Lekki Massacre as 'false'. Understanding the import...

Exploring Controversies in AI-Powered Spam Filters: A Case for Responsible AI

Delve into the seemingly uncontroversial realm of spam filters, where AI plays a crucial role. However, when examined closely, these systems...

Facebook's Struggle to Detect False Information: A Case Study on COVID-19 and Voting Misinformation

This incident underscores the challenges of maintaining trustworthy AI in social media platforms, especially during critical times like elec...

Robot's Unplanned Declaration Sparks Debate on Safe and Secure AI: A Case Study for Responsible AI Governance

This AI incident, where a robot inadvertently declared its potential to 'destroy humankind', underscores the importance of trustworthy AI. I...

Government's Inadequate Approach to Algorithm Usage - A Case Study

This AI incident highlights the need for responsible AI governance. The Government's handling of algorithms in this case falls short, demons...

Unfair Bias in AI: UK Passport Photo Checker Discriminates Against Dark-Skinned Women

This AI incident underscores the importance of trustworthy AI. The passport photo checker, used by UK authorities, exhibited bias against da...

AI Incident Involving Jewish Baby Stroller Image Recognition

An AI incident occurred involving an image recognition system misclassifying a stroller as a gun, raising safety concerns for the Jewish com...

YouTube's Radicalization Trap: The Christchurch Shooter Incident - A Case Study in AI Harm Prevention

Explore the tragic incident involving the Christchurch shooter, revealing YouTube's potential role in radicalization. This AI incident maps...

Apology for Exclusion in Coronavirus Vaccine Distribution Plan: A Case Study on AI Governance

Stanford University faced criticism over a coronavirus vaccine plan that prioritized wealthy individuals, excluding many front-line healthca...

Examining Gender Bias Allegations in Apple Card: A Cautionary Tale for Responsible Fintech AI

This AI incident, involving gender bias complaints against the Apple Card, underscores the importance of trustworthy AI governance. It serve...

Facebook Alleged to Enable Housing Discrimination: A Case for Responsible AI

A case has been brought against Facebook by the U.S. Department of Housing and Urban Development (HUD). The charge is that the social media...

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.