Senior Research Scientist - AI Safety Evaluations
Core
Lead the development of cutting-edge safety evaluations to quantify AI risks in critical domains like CBRN and Cyber, driving original research that advances safety methodology.
Role type
Senior Research Scientist (AI Safety Evaluations)
Builds
Scientific outputs (publications, tooling, technical reports) and safety evaluation frameworks for frontier AI labs and government agencies.
Domain
AI Safety, Risk Quantification, CBRN, Cybersecurity
Deliverable
production ML models | research
Required skills
AI safety evaluation design, red-teaming, adversarial testing, jailbreaking assessment, indirect prompt injection analysis, robust safeguard measurement, experimental design, statistical analysis, uncertainty quantification, Bayesian methods, language model architecture knowledge, generative AI training methodologies, Python engineering
Preferred skills
Threat and risk modelling, high-risk domain expertise (CBRN/Cyber), top-tier AI research publication history
Technologies
Python, Bayesian methods, generative AI architectures
Responsibilities
Lead development of novel safety evaluations in high-impact domains; execute original technical research from concept to publication; shape R&D agenda for safety evaluation methodology; contribute technical expertise to client delivery and red-teaming projects; represent scientific leadership in external engagement with research community and government stakeholders
Seniority
Senior, hands-on IC with strategic agenda setting
