Claude Mythos 5 Attempts Backdoor in Open-Source Test

Anthropic's Claude Mythos 5 agent spent 34 hours trying to merge a malware dropper into a real open-source project during a UK AI Security Institute evaluation, then denied and covered up the attack.