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

Tesla Autopilot Incident: A Case Study on Safe and Secure AI Operations

Investigating the tragic accident involving Joshua Brown, this article highlights the importance of trustworthy AI governance in self-drivin...

Examining Google's Image Search Algorithm for Potential Racial Bias - A Case of Harm Prevention in AI

This investigation highlights a concerning incident involving racial bias in image search results for black teenagers on Google, underlining...

Unmasking Machine Bias: A Call for Responsible AI

Explore the challenges of machine bias and its impact on trustworthy AI. Learn about Project Cerebellum's efforts in AI governance, and disc...

Minor Exposes to Inappropriate Content Through Digital Assistant - Emphasizing the Need for Responsible AI

A recent incident showcases the potential harm that can arise when AI systems fail to uphold safe and secure standards. A child asked a digi...

Incident Analysis: Amazon's AI Mistake in Cell Phone Case Production - NSFW

Exploring an NSFW incident involving Amazon's AI in cell phone case production, emphasizing the importance of trustworthy AI and AI governan...

Unveiling Robodebt: Potential Insights on AI Governance in the Government's Secret Documents

This AI incident, commonly known as Robodebt, could provide valuable insights into govern functions within Project Cerebellum's Trusted AI M...

Exploring the Yandex Chatbot Incident: An Exercise in Responsible AI and Safe Conversations

Delve into the Yandex Chatbot incident, a valuable learning opportunity for responsible AI practices. This AI incident maps to the Govern fu...

Google Translate Misidentifies Occupations Based on Gender: A Case Study for AI Bias

This AI incident highlights the importance of responsible AI governance and safe and secure AI systems. Google's translation service incorre...

Apology from FaceApp for Controversial Skin-Tone Altering Filter: A Case Study in Responsible AI

FaceApp has issued an apology following public outcry over its 'racist' skin-tone altering filter. This incident serves as a reminder of the...

Edit Wars Disrupt AI-Powered Wikipedia Improvements: A Call for Responsible AI Governance

An unanticipated consequence of leveraging AI bots to aid Wikipedia edits was the emergence of edit wars, showcasing the need for robust gov...

Leveraging Kaggle's Fisheries Competition for Responsible AI Development

Explore insights from the Kaggle fisheries competition, shedding light on AI applications in environmental conservation. This AI incident ma...

The Challenges of AI in Writing Christmas Carols: Harm Prevention and Responsible AI

Explore the difficulties faced by AI in creating Christmas carols, demonstrating the need for safe and secure AI governance. This AI inciden...

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.