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

How Ad Technology Can Harm News Funding: A Case Study on Brand Safety

This case study sheds light on the potential negative impact of brand safety technology on news funding. As we advocate for trustworthy AI,...

Israeli Farmer vs. AI Algorithm: A Case Study in Responsible AI Governance

This incident involving an Israeli farmer's dispute with an algorithm highlights the importance of trustworthy AI and robust AI governance m...

Privacy Concerns Prompt University of Illinois to Halt Use of Remote-Testing Software

The University of Illinois has decided to discontinue the use of its remote-testing software following complaints from students about privac...

Examining E-commerce Algorithms for the Prevention of Vaccine Misinformation - A Responsible AI Approach

This AI incident highlights the need for trustworthy AI in e-commerce platforms, focusing on algorithmic curation. By auditing these systems...

Exam Monitoring Software Disadvantages Impact BIPOC Students at University of Toronto - Safe and Secure AI Harm Prevention

An investigation reveals disadvantages faced by Black, Indigenous, and People of Color (BIPOC) students in their exams due to monitoring sof...

Incident Analysis: Potential Use of Music as a Privacy Shield in Livestream Scenarios - AI Governance Implications

This intriguing incident raises questions about the use of music as a privacy measure in livestream scenarios. It maps to the Govern functio...

Facebook's AI-Powered Content Moderation: Analyzing the Rejection of Fashion Ads

Exploring Facebook's AI-driven content moderation systems, we delve into the reasons behind the rejection of certain fashion ads. This AI in...

Addressing Challenges in Combating Inappropriate Content on Social Media Platforms: A Focus on Facebook and Twitter

Exploring the complexities faced by leading social media platforms, such as Facebook and Twitter, in managing content violations in the Balk...

Accidental YouTube Algorithm Blocking of 'Black-and-White Chess Strategy' Highlights Need for Responsible AI Governance

This AI incident, involving YouTube's algorithm mistakenly blocking a popular chess strategy video due to its color scheme, underscores the...

Experience of Using Tesla's 'Full Self-Driving': An Analysis for Safe and Secure AI

This incident provides insights into the practical application of Tesla's Full Self-Driving feature, emphasizing the importance of trustwort...

Exploring Ethical Decision-Making in AI: The Govern Function

Delve into the fascinating world of ethical AI as we discuss the Govern function, a key component of Project Cerebellum's Trusted AI Model (...

Unauthorized Access: A $35 Million Bank Heist Using Voice Cloning - Emphasizing the Need for Robust AI Governance

In an unprecedented cybercrime, fraudsters cloned a company director's voice to execute a $35 million bank heist. This incident underscores...

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.