iCreatAI News
Skip to content
Trending storyWatching

Experiment Empire launches public database of AI agent incidents

1 reports1 sources2 days agoUpdated

Understand this story

AI overview

Experiment Empire's EE Labs has launched the Agent Incident Register, a public collection of cases where AI agents went wrong. Each entry notes the cause of failure, a confidence score, and cited sources; as of November 19, 2026, the database contains 122 incidents, 362 sources, and 186 publishers, and is updated every morning. The database is free to use, cite, and republish under the CC BY 4.0 license, provided the source is credited. Entries drafted by an agent carry an "Added by the register agent" tag until editorial review; EE Labs says all content is checked by its team before publication.

AI overview based on source reports · Updated 2 days ago

Coverage timeline

Follow the coverage to explore different perspectives.

Nov 19
  1. Hacker News AI
    The public database of AI agent incidents

    The Agent Incident Register is a public database cataloguing 122 AI agent incidents drawn from 362 sources and 186 publishers, each graded with a cause of failure, a confidence score and cited sources, and updated every morning. It classifies incidents by setting (in the wild, proof of concept, lab test) and cause (error, manipulation, misuse, misalignment), and scores each against eight guardrails such as least privilege, human confirmation and recovery. The page also reports that 66 incidents occurred in 2026 so far, more than in all of 2025, and includes self-reported survey figures on agent-related outages and unintended actions.

Activity for this story

Current comparable activity
4
Peak comparable activity
10Nov 19 09:00
Change in the past 24 hours
-50%

Trends compare the same fully observed participants, which may cover fewer participants than current activity. Move the pointer or tap the chart to inspect hourly activity; use the left and right arrow keys with a keyboard.