Austin P. Morrissey
AI safety research on biosecurity capability evaluations
I’m a Pivotal Research Fellow working on AI capability evaluations for biosecurity, mentored by SecureBio. My research asks whether large language models meaningfully lower the bar for actors seeking to misuse biology — and how to measure that empirically.
Previously: mRNA scientist (UMass Amherst, Boston University) with wet-lab R&D experience. Now: writing evals that probe the boundary between what models won’t say and what they can’t say.
Read my writing → About me → Research projects →
One does not do research in a cathedral.
Latest writing
No matching items