Meta
Research Scientist, CBRNE AI Safety
$154K–$217K
Menlo Park, CA · Full-time · Posted Oct 22, 2026
- Experience
- 2+ years of research experience in a stated relevant area; the PhD requirement allows equivalent experience in CBRNE safety research.
- Education
- Bachelor's degree in Computer Science, Computer Engineering, a relevant technical field, or equivalent practical experience; PhD in a stated relevant field or equivalent experience in CBRNE safety research.
- Visa sponsorship
- Candidates must be a U.S. citizen or national, U.S. permanent resident (current Green Card holder), or lawfully admitted into the U.S. as a refugee or granted asylum.
About the job
Conduct research to evaluate and reduce AI risks in CBRNE domains, developing safety evaluations, mitigations, and preparedness frameworks.
What you'll do
- Research AI model safety in CBRNE domains and develop evaluation methodologies and benchmarks.
- Design safety mitigations and guardrails; contribute to preparedness frameworks for emerging risks.
- Collaborate with policy, legal, product, and external partners; publish and present research findings.
What they ask for
- Bachelor's degree in Computer Science, Computer Engineering, a relevant technical field, or equivalent practical experience.
- PhD in biology, chemistry, nuclear science, security studies, or a related field, or equivalent experience in CBRNE safety research.
- 2+ years of research experience related to biosecurity, chemical safety, nuclear safety, or weapon non-proliferation.
- Experience with risk assessment and safety evaluation frameworks; track record of published research or contributions in CBRNE safety, AI safety, or related domains.
Nice to have
- Knowledge of dual-use research concerns and responsible disclosure practices.
- Background in AI red-teaming, adversarial testing, or safety evaluation; experience developing safety guidelines or policies.
Pay
$154,000/year to $217,000/year (base salary)
Tools
AI/ML systems, large language models
AI focus
Evaluate and mitigate CBRNE-related risks in AI systems; develop safety frameworks, benchmarks, and guardrails, and assess model capabilities and vulnerabilities.
Model-written summary · the original posting is authoritative