Research Scientist, Manipulation Evals

Apart Research is looking for a Research Scientist for Manipulation Evals to design, build, and publish novel evaluations of AI harmful manipulation that directly inform EU AI Act enforcement. You will own the full pipeline of developing risk scenarios, building evaluations, and running them against frontier models to provide credible evidence for regulatory action.

• Develop and prioritize risk scenarios, then build and run evals end to end against frontier models using coding agents (Claude Code, Codex)
• Design novel evaluation methods for AI manipulation, turning risk scenarios into rigorous, defensible measurements
• Perform literature reviews and track the state of the art in manipulation, deception, and persuasion evals
• Investigate model behavior and red-team your results; scan large volumes of transcripts to find manipulation patterns
• Turn qualitative observations into quantitative measures and red-team your own and others' evals
• Deliver new results at least weekly and set research directions with significant autonomy
• Communicate technical results so regulators can use them; write findings clearly for the EU AI Office and to academic quality standards
• Co-author public deliverables (papers, benchmark releases, technical reports) and engage directly with the Office

• Strong research background in AI safety, evaluation methodology, or related fields
• Experience designing and implementing evaluations or benchmarks
• Proficiency in coding and working with frontier AI models
• Ability to translate technical research into clear communication for regulatory audiences
• Experience with literature reviews and tracking state-of-the-art research
• Strong analytical and red-teaming skills
• Ability to work autonomously and set research directions
• Overlap with EU time zones required; strong preference for Europe-based candidates

Apply via the application form linked on the job page (about 5 minutes).