Researcher, Safety - Frontier Assurance

What you’d do

Build technical foundations for assessing frontier AI risks. Design safety assessments with external evaluators and internal teams, define evidence and methods, and run experiments where direct access isn't possible. Full-time in San Francisco. Found on the employer's careers page on Oct 18, 2026. Apply through the link.

You’ll need

• Design technical AI safety assessments
• Work with external evaluators and auditors
• Define evidence and methods for safety claims
• Conduct experiments and analyses on evaluators' behalf