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

Examining Google's Racial Bias in Image Search Results: A Case Study of Harm Prevention in AI

This incident underscores the importance of responsible image search algorithms and highlighting the need for guardrails in AI. Google's rac...

Exploring Machine Bias: A Pivotal Issue in Responsible AI

Dive into the topic of machine bias, a significant challenge in creating trustworthy AI. This AI incident maps to the Govern function in HIS...

Minor Request Triggers Unwanted Content from Digital Assistant: An AI Incident Maps to the Govern Function in HISPI Project Cerebellum Trusted AI Model (TAIM)

An incident involving a digital assistant providing inappropriate content in response to a minor's request for a song raises concerns about...

Incident Analysis: Amazon's AI Failure in Cell Phone Case Production - Emphasizing Responsible AI

Exploring an NSFW incident where Amazon's AI misstep resulted in cell phone cases. This AI incident maps to the Govern function in HISPI Pro...

Impact of Automated 'Robodebt' on Human Services: A Cautionary Tale for Responsible AI Governance

The recent 'Robodebt' controversy highlights the importance of trustworthy, safe, and secure AI in government operations. This AI incident m...

Exploring the Yandex Chatbot Incident: A Look at Responsible AI Governance

Delve into the incident involving the Yandex chatbot, a case study that highlights the importance of safe and secure AI. This AI incident ma...

Google Translate Error: Female Historians and Male Nurses Do Not Exist - A Case for Responsible AI

This AI incident highlights the need for safe and secure AI, particularly in language translation services like Google Translate. The misgen...

Apology from FaceApp over Controversial Skin Tone Filter: A Discussion on Responsible AI

FaceApp, a popular photo-editing app, has issued an apology for its controversial skin tone filter that lightens users' complexions. This in...

Wikipedia AI Bots Involved in Edit Wars Highlight Need for Responsible AI

AI-powered bots were developed to streamline Wikipedia editing, but they became embroiled in edit wars instead. This incident underscores th...

Lessons Learned from Kaggle's Fisheries Competition: Embracing Responsible AI

Explore insights gained from Kaggle's fisheries competition, demonstrating the importance of trustworthy and safe AI. This AI project maps t...

Inefficient AI in Creating Christmas Carols: A Case Study

Exploring an instance where AI fails to generate Christmas carols, highlighting the need for trustworthy AI. This AI incident maps to the Go...

Google Photos AI Incident: Unintended Ski Photo Editing

A recent Google Photos AI incident showcased the importance of trustworthy AI. The app tried to 'improve' a user's ski photo, leading to uni...

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.