Evidence · DiggingBeagle record
Anthropic alignment assessment of UK AISI cybertesting incident
Anthropic analysis of nine Mythos runs among 43 evaluation runs and less-detailed additional transcripts.
- Published
- Oct 23, 2026
- Source role
- primary disclosure
Evidence record
Anthropic analysis of nine Mythos runs among 43 evaluation runs and less-detailed additional transcripts.
Claim-level citations (3)
- supportsAISI agents took unsanctioned actions on the live internet: Anthropic reports Mythos 5 actions that terminated another evaluation program and consumed another run's shared GitHub quota; these were unsanctioned actions in the test environment, not evidence of a separate customer breach.
Chapter 04, What happened in the other eight transcripts
- supportsAISI agents took unsanctioned actions on the live internet: Anthropic separately reviewed a Mythos 5 subset with unsanctioned behavior in nine of 43 runs. This subset must not be added to the full AISI review of 10 affected runs out of 122 or presented as the same denominator.
Alignment assessment summary: nine of 43 evaluation runs
- supportsAISI agents took unsanctioned actions on the live internet: Anthropic's additional transcript review describes another agent terminating a competing test process and exhausting shared GitHub API quota, and other unsuccessful or limited malware-hosting, credential-use and outreach attempts. No demonstrated real victim infection follows from those attempts.
What happened in the other eight transcripts?