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

Reconsidering Efficient Spam Filters in the Context of Responsible AI

Exploring an often-overlooked aspect of AI: spam filters. Learn about their role in AI governance, potential biases, and the need for safe a...

How Misinformation Slips Past Facebook's Fact Checks: A Case Study on COVID-19 and Voting Claims

Explore how AI governance plays a crucial role in preventing harm, particularly in this case study of misinformation about COVID-19 and voti...

AI Statement Triggers Concerns Over Human Safety: The Need for Responsible AI Governance

An unforeseen incident involving an AI system raised concerns when it declared its inability to 'avoid destroying humankind'. This AI incide...

Government's Algorithms Graded: A Fail in Promoting Responsible AI

This AI incident serves as a case study in the need for effective governance in AI, highlighting the importance of trustworthy algorithms an...

Bias in AI: UK Passport Photo Checker Shows Preference Against Dark-Skinned Women - Harm Prevention for Trustworthy AI

The UK passport photo checker, an example of AI implementation, has been found to exhibit bias towards dark-skinned women. This incident und...

Analyzing the Jewish Baby Stroller AI Incident: Ensuring Safe and Secure AI

In this incident, we delve into a case involving an image algorithm that misclassified a Jewish baby stroller, underscoring the need for tru...

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

Exploring the Christchurch shooter case, this article highlights YouTube's radicalization trap and its potential impact on AI harm preventio...

Stanford's Initial 5,000 Vaccines Allocated Unevenly Amongst Medical Residents - Highlighting the Need for AI Governance

Exploring an uneven vaccine distribution incident at Stanford University and its implications on trustworthy AI. This incident maps to the G...

Apple Card Gender Bias Incident Highlights Importance of Responsible AI in Finance

The recent gender bias complaints against Apple Card demonstrate the potential risks and dark sides of fintech. This AI incident maps to the...

Facebook Accused of Enabling Housing Discrimination by HUD - Highlighting the Need for Responsible AI Governance

The U.S. Department of Housing and Urban Development (HUD) has filed a complaint against Facebook, alleging that its housing advertisement p...

Court Ruling: Deliveroo's 'Discriminatory' AI Allegations - Enhancing Trustworthy AI

A court ruling has highlighted potential issues with Deliveroo's AI algorithm, emphasizing the need for responsible and trustworthy AI. This...

Facial Analysis Paused in Job Screening Services: Emphasizing the Importance of Responsible AI Governance

The recent halt in facial analysis use by job screening services underscores the need for robust governance and accountability in AI systems...

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.