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

Addressing Bias in AI: The Islamophobia Issue

Exploring an unfortunate reality - the biases in AI that can contribute to Islamophobia. This AI incident maps to the Govern function in HIS...

Controversial Decision by Xsolla Leadership: Layoffs Using Big Data and AI Analysis

Incident involving job losses at Xsollai mapped to the Govern function in HISPI Project Cerebellum Trusted AI Model (TAIM). CEO's letter add...

Uncontrolled GPT-3 Bot Interaction on Reddit: A Case Study in the Need for Responsible AI Governance

This AI incident serves as a cautionary tale, highlighting the importance of robust governance for safe and secure AI. The unsupervised inte...

Potential Use of Autonomous Weapons in Libya: A Case for AI Governance

This incident raises questions about the role of AI in modern warfare, particularly autonomous weapons. It highlights the need for robust AI...

Settlement of $550 Million Over Facial Recognition Tech Highlights Privacy Concerns in AI

This AI incident involving Facebook's facial recognition technology underscores the necessity for privacy protection in the development and...

Misleading AI Promises in Healthcare: A Case Study for Responsible AI

Exploring an instance of misleading AI claims in the medical field, this article underscores the need for trustworthy AI and safe & secure A...

Algorithm Bias Exposure in Healthcare: An Unfair Distribution of Care to Black Patients

Uncovering a concerning case of algorithmic bias, this incident illustrates how AI can perpetuate racial disparities in healthcare delivery....

California's New Law Targets Amazon Warehouses: A Step Towards Safe and Secure AI Workplaces

The new California law aims to improve warehouse conditions for workers, emphasizing the need for trustworthy AI in work environments. This...

Unintentional Fire Incident at UK's First Online-Only Grocer: A Case for Safe and Secure AI

An unexpected collision between robots ignited a fire at an online-only grocery store in the UK, raising concerns about the need for reliabl...

Microsoft Automates Journalism Role: A Case Study for Responsible AI

In this analysis, we delve into Microsoft's decision to replace human journalists with robots. This move highlights the need for trustworthy...

Ethical Guardrails for AI: Tech Companies Push for Responsible AI Governance

Explore the recent moves by tech firms to establish ethical guardrails around AI development, contributing to the growth of trustworthy and...

Mass Shooting Videos vs Car Washes: An Unfortunate Mishap in Facebook's AI Moderation System - Highlighting the Need for Responsible AI Governance

Incident highlights a lapse in Facebook's AI moderation system, where harmless videos of car washes were incorrectly flagged as violent cont...

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.