AI Risk Atlas Prototype/DemoUnofficial independent experiment. Not an official xAI product. Scores can be wrong.

Back to watch register
15Human agencyBelow the top 20

Benefits eligibility with no real appeal

AffordanceImpact domainBoth

Statement (NASA form)

Given that welfare, disability, and asylum systems already hide behind a model score, there is a possibility of a refusal that a human reviewer only rubber-stamps resulting in a right that exists in law and dies in a queue.

Likelihood
3Probable
Consequence
4Critical
Urgency
3Priority

An appeal against a score the appellant cannot see is not an appeal.

Composite 15 = 3×4 + 3

Applicable mitigations

Related on the map