Researcher: talk directly to AI labs' biosecurity teams, not their CEOs' polished takes

anshulkundaje · x · 2026-09-16

Stanford's Anshul Kundaje argues AI orgs are not monocultures — there is substantial internal debate on safety issues and severity. He recommends critics engage directly with company biosecurity teams, which is a very different experience from polished blog posts and dumbed-down CEO interviews.

Related event: Stanford researcher says nearer AI biosecurity risks lie in supply chain, not AI-designed pathogens(6 posts)→

Original post →

More from AGI Musings

AGI Musings channel →