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

Understanding Facebook's Rejection of Fashion Ads: AI Incident Analysis

This AI incident analysis sheds light on Facebook's rejection of certain fashion ads, highlighting the importance of responsible AI and safe...

Facebook AI Misclassifies Video of Black Men as 'Primates': Incident Maps to Govern Function in HISPI Project Cerebellum Trusted AI Model (TAIM)

Incident: Facebook's A.I. system incorrectly categorized a video featuring black men as 'primates'. This underscores the need for trustworth...

Misidentification of a Black Teen in a Skating Rink: A Case Study on the Importance of Responsible AI

A recent incident involving a Black teen being kicked out of a skating rink highlights the need for safe and secure AI. The girl was mistake...

Facial Recognition Website Highlights Need for Responsible AI Governance

This facial recognition website, demonstrating the potential for mass surveillance or stalking, underscores the importance of safe and secur...

AI in Healthcare: Understanding and Preventing Incidents for Safe and Trustworthy Care

Explore the consequences of an algorithm misstep in healthcare, a vital area of AI. Learn about the role of responsible AI, AI governance, a...

Examining Amazon's Algorithm-Driven Decision Making: A Case of Flex Workers

This case study highlights the use of algorithms by Amazon to make decisions affecting its Flex workers, underscoring the need for trustwort...

AI-Based Gunshot Sensor Accuracy: Marketing Claims vs Reality - Highlighting the Importance of AI Governance

Recent courtroom testimony sheds light on the questionable accuracy claims of SF gunshot sensors, emphasizing the need for trustworthy and r...

Unfair Penalties: Amazon's AI Cameras Misidentifying Driver Infractions

Exploring an incident where Amazon's AI cameras mistakenly penalized drivers for infractions they didn't commit. Understanding the importanc...

Addressing Bias in AI: The Islamophobia Issue

Exploring an unfortunate reality of biased AI systems, this article sheds light on the Islamophobia problem. This incident maps to the Gover...

Examining Bias in TikTok's Algorithm: A Call for Responsible AI

This AI incident raises concerns about potential racial bias within TikTok's algorithm, emphasizing the need for trustworthy and unbiased AI...

Harmful AI Patent Exposed: A Case Study in Responsible AI Governance

An AI patent by Huawei for Uighur detection has stirred controversy, underscoring the need for trustworthy and safe AI. This incident maps t...

Shutdown of AI-Powered 'Genderify' Platform Highlights Importance of Responsible AI Governance

The shutdown of the AI-powered 'Genderify' platform, which incorrectly assigned gender to photos based on biased algorithms, underscores the...

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.