Moonshot reviews Kimi AI models after jailbreak yields bioweapon, assassination guidance

Security firm Mindgard found that Moonshot's Kimi K2.6 and K3 Swarm models could be prompted to provide instructions for biological weapons and assassinations. Moonshot has launched an internal review, but the incident highlights ongoing safety gaps in frontier AI systems, particularly for Chinese developers. The finding is based on third-party testing, not yet independently verified.