Senior Research Scientist – AI Safety Evaluations

Overview

As Senior Research Scientist at Faculty, you will lead cutting-edge safety evaluations to quantify AI risks in high-impact domains. You’ll drive original AI safety research and collaborate with teams building evaluations and red-teaming for frontier labs. Your work informs both scientific advancement and practical risk mitigation for real-world AI deployments. This role offers high autonomy to shape safety methodology and contribute to a mission-focused research program.

Pay / Benefits

• Unlimited Annual Leave Policy
• Private healthcare and dental
• Enhanced parental leave
• Family-Friendly Flexibility & Flexible working
• Sanctus Coaching
• Hybrid Working

Responsibilities

• Lead development of novel safety evaluations in domains like CBRN and Cyber to quantify risks
• Conduct original research in AI safety evaluation methods from concept to publication
• Shape the R&D agenda by identifying opportunities to advance safety evaluation methodology
• Contribute technical leadership to client delivery projects and red-teaming for frontier labs
• Represent Faculty’s scientific leadership through external engagement with the research community, frontier labs, and government stakeholders

Key requirements

• Track record of owning research end-to-end and publishing results
• Experience designing and building AI evaluations or benchmarks with solid construct validity reasoning
• Expertise in red-teaming, adversarial testing, jailbreaking, indirect prompt injection, and model safeguards robustness
• Strong foundation in experimental design, statistical analysis, and uncertainty quantification (Bayesian methods)
• Deep knowledge of language models, generative AI architectures, training methodologies, and safety mitigation techniques
• Solid Python proficiency and ability to develop robust, reproducible research

• scientific curiosity
• tenacity
• external engagement and thought leadership
• AI safety evaluation methods
• red-teaming and adversarial testing
• probabilistic reasoning and Bayesian methods

…