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

Unauthorized Use of Alexa Leads to Police Raid: A Case Study in AI Misuse

This AI incident highlights the potential risks of unregulated AI use, specifically voice assistants like Amazon's Alexa. In this instance,...

Potential Discovery of Commercial Filtering Mechanism in Amazon Echo Devices

This intriguing discovery suggests a possible mechanism employed by Amazon Echo devices to prevent activation during commercials, contributi...

Misunderstood Algorithmic Termination: Ensuring Safe and Secure AI

Incident involving an inexplicable algorithm-driven dismissal raises concerns about trustworthy AI governance. The HISPI Project Cerebellum'...

Unintended Outcome of Chinese AI in Bus Advertisement Highlights Importance of Safe and Secure AI

An AI system misidentified a passenger on a bus in China, triggering a call to the police. This AI incident maps to the Govern function in H...

Assessing Hiring Algorithms for Bias: An Examination of AI Governance

This article discusses the challenge of auditing hiring algorithms for bias, emphasizing the importance of AI governance for ensuring fairne...

Tragic Self-Driving Uber Accident in AI-Friendly Arizona Highlights Need for Responsible AI Governance

This AI incident involving a self-driving Uber vehicle underscores the importance of trustworthy and safe AI. It maps to the Govern function...

Elite: Dangerous AI Incident: Unintended Superweapon Development Demonstrates Importance of Responsible AI

In the popular space simulation game Elite: Dangerous, an incident occurred where unforeseen circumstances led the in-game AI to develop sup...

AI-generated Fake Speech by Former President Obama Highlights Need for Responsible AI

This incident showcases an example of AI misuse in creating a fake speech attributed to former U.S. President Barack Obama. Such incidents u...

The Role of AI in Criminal Justice: Preventing Wrongful Convictions

This AI incident highlights the importance of responsible AI governance in criminal justice systems. It reveals potential flaws in decision-...

Violent Content Feeding Leads to Unintended AI Behavior: A Case Study on MIT's AI Experiment

Exploring an incident where MIT scientists inadvertently created an aggressive AI by exposing it to violent content from Reddit, this articl...

Exploring AI in Labor Markets: The Case of the National Residency Matching Program

Investigating the role of AI in the National Residency Matching Program as a labor market solution. This AI incident maps to the Govern func...

Examining Racial Bias in AI Systems: A Step Towards Trustworthy AI

Investigating a recent incident involving racial bias in an AI system, we delve into the importance of safeguarding our technology from harm...

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.