Frontline Evaluator - Frontier AI Systems

Highlights

• Investigate honesty and alignment of frontier AI systems in the wild.
• Run rapid-response investigations of major AI incidents for public reports.
• Serve as independent evaluator for governments including the EU.

Ideal candidate

Empirically sharp investigator who turns ambiguous AI behavior concerns into rigorous tests and clear public-interest reports. Resourceful, collaborative, and motivated to make frontier AI oversight matter.

Requirements

• Strong empirical judgment to extract findings from data and identify anomalies
• Proficiency in Python for data analysis and evaluation tooling
• Ability to turn an ambiguous concern into a testable question
• Operational resourcefulness and sound prioritization with incomplete information
• Ability to iterate quickly balancing scrappiness and thoroughness
• Strong communication skills explaining technical findings to varied audiences
• Collaborative, low-ego orientation open to giving and receiving feedback