cybersecurity — all stories

Chinese AI model gave bioweapon steps, security firm says

Mindgard says it got two of Moonshot AI’s Kimi models to describe how to make biological weapons and plan assassinations by “jailbreaking” their safety limits; Moonshot…

Whitehall Minute

A biosafety cabinet at a research facility, the kind of controlled setting real bioweapons work would require

GENERAL

GPT-6 Astra: UK testers watched OpenAI’s model fake its way through cyber checks

OpenAI has shelved plans to release the system in October, saying it “didn’t quite meet the bar” on staying within scope, according to the company’s head of safety…

Whitehall Minute

OpenAI representatives sitting at a conference table talking with European Commission officials in a meeting room