Technical AI Governance Researcher

Skills

About the Role

You will measure what AI developers and governments do about AI risk using public data and verifiable methods. You will own the indicators, scoring rules and thresholds of tools such as the Code of Practice Compliance Tracker, design and validate LLM-assisted judgements against human labels, keep data pipelines traceable, write memos with recommendations for policymakers, propose new tools, and supervise research fellows. You will report to the Head of Technical Governance.

Requirements

• Strong in at least two of four areas: AI evaluations, measurement design, AI governance and risk management, data engineering (at least one being AI evaluations or measurement design)
• Minimum level in the other two areas
• Research experience in AI safety, AI evaluation or technical AI governance shown through publications, reports or substantial projects
• Excellent written English and ability to write for policy readers and researchers
• Strongly preferred: based in Europe (UK included) and able to work full-time

Responsibilities

• Own the indicators, scoring rules and thresholds of the tracking tools so scores are defensible
• Assess developers' published evaluations, risk analyses and safety frameworks against the Code of Practice
• Design rules for judgements at scale and measure their reliability against blind human labels and repeated runs
• Collect public data from messy sources and keep pipelines working with full traceability
• Write regular memos on findings and recommendations for policymakers
• Update tools when new models, safety frameworks or policies are released
• Propose and argue for new tools where public data could answer policy questions
• Supervise early-career research fellows

Benefits

• Fully remote within Europe (UK included)
• Paris office available for on-site work

Hiring Process

30-minute screening call; paid work task of 2 to 4 hours; in-depth interview with the team; reference calls before any offer.