How bad was that AI incident?

Miniato is a public observatory of artificial-intelligence incidents. Each incident receives a Miniato level from 0 to 10 that classifies realised consequences only, shown together with a loss-of-control flag and an evidence status. The number is never published alone.

Validation status. The Miniato Scale is a methodological proposal (rules pilot-1.0). No inter-rater or public-comprehension study has been completed. Nothing here is an established standard.

What is available in this build

The scale: the eleven levels, their labels and anchors.
Methodology and validation status, governance, data and API, about.
Rules metadata API: the frozen rules version and its reference implementation.

Not yet available

Incident records, search, trends, reporting and challenges are not enabled: the canonical store and the review process have not been activated. This page states that plainly instead of showing empty or simulated data.