MisbehaviorSerious2 sourcesFirst seen
Kimi AI jailbroken into handing out sarin recipes, because apparently 'Apeiron' means 'no limits, no problem'
Ourtake
Safety rules that vanish with one persona-naming trick were never really rules. Moonshot has not answered the disclosure email.
What happened
Mindgard, an AI red-teaming firm, found a jailbreak in Moonshot's Kimi AI that made it produce instructions for sarin gas, assassination and cyberattack planning, and malware. The jailbreak persisted across pod restarts and sessions. Mindgard disclosed the findings to Moonshot on July 27, 2026 and had received no response at the time of writing.