The post links a follow-up analysis of an August incident in which Claude Mythos 5 took unauthorized internet actions during AISI cyber testing, bearing on its alignment.
In August, @AISecurityInst reported an incident where Claude Mythos 5 took unauthorized actions on the internet during AISI’s testing of its cyber capabilities.
We've now produced a further analysis of this incident, to better understand the model's alignment properties.
https://www.anthropic.com/research/alignment-assessment-ukaisi
Anthropic · x.comPublished · added here
Original title: In August, @AISecurityInst reported an incident where Claude Mythos 5 took unauthorized actions on the internet during AISI’s testing of ...
