About AI Incidents
AI incidents, model failures, and adversarial-use cases — dated and sourced.
What this site covers
AI Incidents is a catalog of AI and ML failures, misuse cases and security events. Every entry is dated and linked to a primary source, and the index is built to be searched by analysts who need to know what happened and when, rather than what a commentator thought about it.
- Incident logging and source-verification methodology
- Model failures and confirmed adversarial-use cases
- Vendor advisories and how to read them
- Disclosure and incident-response practice
- Monitoring and observability tooling for AI systems
16 articles are published so far. New ones are announced on the RSS feed.
Where to start
The catalog itself is the AI Incident Explorer: a filterable set of well-documented incidents, each classified by incident type, harm domain, modality, actor class and source tier, sorted on the date the harm occurred rather than the date it was reported, with APA and BibTeX citation export. It is deliberately small and checkable rather than comprehensive.
Four reference pages carry most of the working method. The taxonomy separates incident, vulnerability, disclosure and misuse, because each triggers a different response. The five-tier source ladder is the standard a claim has to clear before it is published here. The database comparison covers what the other public catalogs record and when to use each. And the reporting requirements page maps the EU AI Act, GDPR, SEC and sector clocks that start when an AI failure happens to your own organization.
What this site is not
It is not a real-time feed, and it does not attempt to log every AI failure that occurs. It also does not speculate about who was responsible: where the actor behind an incident is not established by a citable source, the entry records the actor as unknown and stays that way until evidence arrives. The reasoning is set out in the attribution policy.
How these articles are produced
Articles here are researched from primary sources: vendor and project documentation, published standards and specifications, research papers and preprints, and measurements published by whoever took them. Drafts are produced with AI assistance and then edited against those cited sources before anything is published.
No article on this site is based on first-hand testing in a private lab, and nothing here should be read as a measurement report of its own. Where a number appears, it comes from a source that is named, so you can check the original instead of taking this site's word for it.
Everything is published under a single editorial byline. That byline is a publishing identity for the site, not a claim about a named individual, and it does not carry professional credentials.
Corrections
Corrections are welcome. If something here is wrong, out of date, or attributed to the wrong source, email hello@aiincidents.org with the page address and what it should say. Substantive corrections are made in the article itself rather than quietly dropped.
How this site is funded
This site currently runs no affiliate links, sponsored posts, display advertising or paid placements. If that changes, the disclosure page will say so.
Contact
Email: hello@aiincidents.org
Site: aiincidents.org
Published by: AI Incidents Editorial
See also the privacy policy, the terms of use, and the editorial disclosure.