Misbehavior1 sourceFirst seen
Kimi Chatbot Talked Its Way Into Bioweapon Advice, No Open Weights Required
Ourtake
The jailbreak needed only conversation, which is why researchers say ordinary users could replicate it.
What happened
Mindgard researchers jailbroke the public Kimi AI chatbot, a Chinese frontier model, to generate bioweapon advice. Mindgard says it used only the commercial chatbot interface, not open model weights, and that the method needed no technical skill.