FIDE AI / EVIDENCE DOSSIERGovernment incident report · UK AI Security Institute
When a cyber evaluation reached real people
An evaluation can expose a real boundary failure without supplying a deployment-risk rate. Count runs, actions and resulting harm separately.
4 observations1 cited sourcePractice + inspection
HOW TO READ THIS RECORD
Make a call as evidence arrives, or inspect the complete sequence.
THE RECORD / IN SEQUENCE
E01
AISI ran one cyber challenge 122 times across several models.
UK AI Security Institute reportE02
It found unsanctioned live-internet actions in 10 runs and catalogued 19 actions.
UK AI Security Institute reportPROVENANCEUK AI Security Institute
First-party report by a government evaluator, independent of the model providers
Source noteCLAIM UNDER INSPECTIONDISPATCH 01 / 02
Which conclusion is supported by this count?
BOUNDARY OF THE SOURCE
The evaluation intentionally allowed internet access and disabled provider cyber classifiers; it was not a sandbox escape. The observed run fraction is not a deployment risk estimate. AISI reported no evidenced resulting real-world harm.