Science and health

AI biosecurity and scientific misuse assessment

AI biosecurity and scientific misuse assessment is the work of testing what AI systems can do in the life sciences, and where that could cause harm. It happens under controlled conditions, so AI companies can set safeguards before release.

As models get better at scientific reasoning, AI companies are being asked to show they have checked for biological misuse. Most teams have no one who can do that. The hires are tricky because the work sits between biosecurity, biosafety, life sciences and model evaluation, and few people have all four. Judgement and discretion matter as much as expertise. Authorised access and clear handling rules are essential, and a strong candidate will ask about them first.

Emerging 4

Roles in this family

4 roles

Where the people come from

Biosecurity, biosafety, life sciences or authorised model evaluation

Not to be confused with. This family describes authorised assessment and safeguards.

Core skills

  • AI red teaming
  • Model evaluation
  • AI risk management
  • Scientific domain expertise

Where it is used

  • Drug discovery and biotechnology
  • Defence and national security
  • Frontier AI governance

More in Science and health

← All new job families

Start a search

Hiring in aI biosecurity and scientific misuse assessment?

Tell us the work you need done. We find the people already doing it, whatever their job title says.