Struggles Against AI-Related Content Violations: A Look at Facebook and Twitter
Facebook, Twitter are grappling with AI challenges in moderating Balkan content, highlighting the need for robust AI governance and harm pre...
Read moreValues 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
Data source & citationAccidental YouTube Algorithm Blockage of the Classic Black-and-White Chess Strategy: A Case Study in Safe AI
An examination of an incident where YouTube's algorithm unintentionally censored a classic black-and-white chess strategy video, highlightin...
Read moreMisinterpretation of Visual Cues in Tesla's Full Self-Driving Technology: A Case Study on AI Misperception
Exploring an incident where Tesla's Full Self-Driving technology misidentified objects such as the moon, billboards, and Burger King signs....
Read moreDubious Moral Advice from an AI Oracle on Reddit: A Case Study in Responsible AI
Exploring an AI incident where the AI oracle of Delphi offered moral advice based on Reddit posts, highlighting the importance of trustworth...
Read moreUnsettling YouTube Videos Misleading Children - A Harm Prevention Case Study
This incident highlights the need for guardrails in AI, specifically in video recommendation algorithms. The case involves children being ex...
Read moreRevising AI Practices: Software Company Embraces Responsible AI Changes
This software company's AI incident underscores the need for trustworthy AI practices. Join us in promoting safe and secure AI by making a d...
Read moreAnalyzing the COMPAS Recidivism Algorithm: A Step Towards Responsible AI Harm Prevention
Dive into our analysis of the COMPAS recidivism algorithm, a crucial example demonstrating the importance of safe and secure AI. This AI inc...
Read moreBias Mitigation in AI: Debiasing Word Embeddings - A Key Step towards Trustworthy AI
Dive into the essential issue of bias in AI, specifically debiasing word embeddings. Understanding and addressing biases is a crucial compon...
Read moreStudy Reveals Bias and Inflexibility in AI's Civility Detection: A Call for Responsible AI Governance
A recent study uncovers bias and inflexibility issues in AI's civility detection. This AI incident maps to the Govern function in HISPI Proj...
Read moreExploring Google's AI Bias Incident: Homosexuality and Trustworthy AI
This incident highlights a potential bias in Google's AI, expressing opinions about homosexuality that are both inappropriate and unacceptab...
Read moreAmazon's Algorithm Modification for Safeguarding Users from Inappropriate Content: A Case of GLBT Books
Explore the recent Amazon incident where their search results and rankings were manipulated to protect users from explicit content related t...
Read moreApology Issued by Google Over Racially Biased Auto-Tagging Incident in Photo App
Google has recently apologized for a racist auto-tagging incident in their photo application. This AI incident maps to the Govern function i...
Read moreData 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.