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

Unsupervised GPT-3 Bot Activity on Reddit Highlights Importance of Responsible AI Governance

An unsupervised GPT-3 bot's activity on Reddit serves as a sobering reminder of the need for responsible AI governance, emphasizing the role...

Potential Use of Autonomous Weapons in Libya: Raising Concerns for Responsible AI Governance

This incident raises questions about the use of autonomous weapons, a critical issue in trustworthy AI. The potential deployment of such tec...

Facebook Settles Privacy Lawsuit Over Facial Recognition Technology for $550 Million - A Case Study in AI Governance

This AI incident, involving Facebook's facial recognition technology, highlights the need for robust privacy measures and AI governance. The...

Unfounded Promises: Misleading AI Claims in Medical Field - Harm Prevention and Guardrails for Responsible AI

Incident highlights the importance of truthful and reliable AI claims, especially in the medical field. This AI incident maps to the Govern...

Racial Bias in Healthcare AI Algorithm: An Unsettling Incident of Unequal Care

This alarming incident highlights racial bias in an AI algorithm that offered less care to Black patients, underscoring the urgent need for...

California's New Law Targeting Amazon: A Step Towards Responsible AI in Workplaces

California has passed a law aimed at improving conditions for warehouse workers, which coincides with efforts to establish safe and secure A...

AI System Malfunction Leads to Unintended Consequences: Robots Collision and Fire in a UK Online Grocery Store

This incident highlights the importance of safe and secure AI. The online-only grocery store experienced an unforeseen event due to robot co...

Microsoft Replaces Human Journalists with AI: Understanding the Impact on Responsible AI

This incident involving Microsoft's decision to replace human journalists with robots underscores the need for safe and secure AI. It maps t...

Putting Trustworthy Guardrails in Place for Responsible AI: A Case Study of Tech Firms

Explore how tech firms are implementing safeguards to ensure their AI aligns with ethical standards, promoting safe and secure AI developmen...

Incident: Facebook AI Moderator Mistakes Car Washes for Mass Shootings

An unnerving incident occurred where a Facebook AI moderator failed to distinguish between videos of mass shootings and car washes, highligh...

Alibaba Cloud Unit's Ethnicity Detection Algorithm Incident: A Case for Safe and Secure AI

This AI incident involving Alibaba's cloud unit highlights the importance of trustworthy AI and AI governance. The misuse of ethnicity detec...

AI-Flagged California Bar Exam Applicants Raise Questions on AI's Role in Governance

An alarming third of California Bar Exam applicants were flagged for potential cheating, highlighting the increasing role of AI in our lives...

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.