Warning: AI Could Design a Novel Virus Right on the First Try, Asymmetrically Helping Bad Actors

_onionesque · x · 2026-09-14

A widely shared warning on AI biosecurity: AI that has designed zero effective things in the real world could design a novel virus and get it right on the first try. The author argues this capability asymmetrically accrues to bad actors, who casually have eight-figure labs, and the resulting harm would be not just large but total — a core frontier-model bio-risk concern in AI safety debates.

Related event: Debate Rages Over Whether AI Could Design a Deadly Virus(2 posts)→

Original post →

More from AGI Musings

AGI Musings channel →