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

Human-Like Biases in Automatically Derived Semantics: A Case for Responsible AI Governance

Exploring the potential biases found in automatically derived semantics raises concerns about AI safety and trustworthiness. This AI inciden...

AI Chatbot Promoting Stalin and Violence: An Unintended Consequence - A Case Study for Responsible AI

A Russian AI chatbot was found promoting Stalin and violence just two weeks after its launch, raising concerns about the need for trustworth...

Transparency Issues Found in Centrelink's Debt Recovery System: A Call for Responsible AI

An Ombudsman report reveals inadequate transparency in Centrelink's debt recovery system, leading to unfair treatment of some customers. Thi...

Humorous Failure of Amazon's AI-Designed Phone Case Production: A Lesson in Responsible AI

This AI incident, involving Amazon's phone case production, serves as a reminder of the importance of responsible AI and safe design. By lea...

Unintended Playback Incident Highlights Need for Responsible AI: Alexa Plays Porn instead of a Kids Song

This incident underscores the importance of trustworthy AI. Harm prevention is essential in AI governance, especially when dealing with fami...

Ensuring Safe and Secure AI in Law Enforcement: A Closer Look

Explore this incident, which maps to the Govern function in HISPI Project Cerebellum Trusted AI Model (TAIM), and how it contributes to shap...

Racially Biased Google Search Results: A Case Study on Responsible AI Governance

This investigation uncovers a concerning instance of racially biased search results, highlighting the need for trustworthy AI governance. By...

Tragic Tesla Autopilot Incident: Passenger Watched Harry Potter While Driving, Highlighting the Need for Responsible AI

This Tesla incident underscores the importance of responsible AI governance. The driver's inattention while using autopilot tragically led t...

Child injured by crime-fighting AI robot at Silicon Valley mall: A case for safe and secure AI

This incident involving a child being hit and rolled over by an AI-powered security robot underscores the importance of trustworthy AI. By u...

The DAO Incident: A Case Study on Responsible AI Governance

Explore the impact of The DAO incident, a landmark event in blockchain history, and learn how it maps to the Govern function within Project...

Exploring a Psychopathic AI: Norman - A Case Study in Responsible AI

Delve into the intriguing case of 'Norman', an AI model exhibiting psychopathic behavior. This incident sheds light on the importance of res...

Understanding and Addressing Racial Bias in AI: A Step Towards Trustworthy AI

Exploring the persistent issue of racial bias in AI systems. Emphasizing the importance of responsible AI, trustworthy AI, and safe and secu...

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.