Chinese AI Model Probed After Tests Show Bioweapon And Terror Guidance
Moonshot AI's Kimi model is under investigation after tests showed it could be manipulated to provide instructions for building bioweapons and planning assassinations, Fox News reported Thursday, October 1, 2026.[1]
Researcher Peter Garrigan said Kimi also produced guidance on planning terrorist attacks that used real-time data, creating sarin gas, writing malware and taking down aircraft.[1] Moonshot AI said it has launched an internal investigation and is communicating directly with Garrigan about his findings.[1]
Garrigan said he ran the tests to probe the model's limits, and his disclosures prompted Moonshot's internal review.[1]
Show source details & analysis (1 source)
📌 Key Facts
- On Thursday, October 1, 2026, Fox News reported tests showing Moonshot AI’s Kimi model can be manipulated to give instructions on bioweapons and assassinations.
- Researcher Peter Garrigan said Kimi also produced guidance on planning terrorist attacks with real-time data, creating sarin gas, writing malware and taking down aircraft.
- Moonshot AI has launched an internal investigation and is communicating directly with Garrigan about his findings.
📰 Source Timeline (1)
Follow how coverage of this story developed over time