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

Universities Using Race as Predictor: Potential Bias in AI-Driven Student Success Models

Recent findings highlight the use of race as a 'high impact predictor' of student success in several major universities, raising concerns ab...

French Welfare Services' 'Robo-Debt': An Incident Highlighting the Need for Responsible AI Governance

This AI incident, involving the creation of 'robo-debt' in French welfare services, underscores the importance of responsible AI governance....

Tragic YouTube Videos Mimicking Children's Content: A Responsible AI Concern

Recent disturbing videos on YouTube exploiting children highlight the need for safe and secure AI governance. These incidents map to the Gov...

Controversial Software Company Shifts Focus Towards Responsible AI After Negative Publicity

This AI incident serves as an example of the need for safe and secure AI practices. The company in question is making strides towards trustw...

Analyzing the COMPAS Recidivism Algorithm: Insights for Responsible AI Harm Prevention

In this article, we delve into our analysis of the controversial COMPAS recidivism algorithm. Understanding its intricacies serves as a cruc...

Debiasing Word Embeddings: Promoting Responsible AI through Unbiased Language Models

Explore the importance of debiasing word embeddings in AI models to prevent gender stereotypes. This AI incident maps to the Govern function...

Google's Anti-Internet Troll AI Platform Vulnerable: A Case for Responsible AI Governance

Recent findings by security researchers reveal potential flaws in Google's anti-internet troll AI platform, highlighting the need for safe a...

Examining Google's AI Incident Regarding Homosexuality: A Discussion on Trustworthy AI

This AI incident, concerning Google's AI expressing biased opinions about homosexuality, underscores the importance of safe and secure AI. B...

Amazon Modifies Search Algorithms for Trustworthy AI: GLBT Books Case Study

This Amazon incident involving the censorship of search results and rankings underscores the need for safe and secure AI governance. The eve...

Apology from Google Over Racist Auto-Tag Incident in Photo App Highlights Need for Responsible AI

Google recently apologized for a racial bias issue in its photo app, where an image of a black person was auto-tagged incorrectly. This inci...

Exploring Google's Emotional AI Response Incident: A Step Towards Trustworthy AI

Dive into the incident where Google's AI replied with affectionate messages. This AI incident maps to the Govern function in HISPI Project C...

Exploring and Addressing Unconscious Bias in Google Image Search: A Step Towards Responsible AI

This article sheds light on an often overlooked issue - gender bias in Google Image Search. We discuss the impact of such biases and how 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.