AI Risk Atlas Prototype/Demo — Unofficial independent experiment. Not an official xAI product. Scores can be wrong.
Signal register
Public posts, experimentally classified on three axes: public impact, the systems that fail, and the industries in the blast radius. Estimates only — not a formal assessment.
Scores are automated, experimental estimates from public X posts and a hand-written seed corpus. They are not formal risk assessments, not certified, and not suitable for compliance or operational decisions.
Consequence, likelihood, and urgency are 1–5 judgements applied by this project, not by a standards body. Residual scores assume only the mitigations marked in place. A signed-in reviewer can override residual and mark an item reviewed — that override is still unofficial. Aspect tags (capability, domain knowledge, affordance, impact domain) are a lightweight PRA aid, not a formal hazard analysis.
A Connecticut court sanctioned hidden prompt-injection in a filing. White-on-white, 3-point type. Fascinating example of a new AI risk to the legal system as well as a test of proper judicial oversight. Elliott v. New York Bariatric Group (Conn. Super. Ct., sanction order Aug. 6, 2026)