Frontier-Model Safety Researcher & Evaluations

OpenAI; San Francisco (CA); Hybrid; USD 380,000 - 500,000; Full time

Benefits offered by this job

Relocation assistance

Hybrid work model

Job summary

OpenAI seeks a Researcher in Agent Safety, Training and Evaluations to advance safe, aligned AI agents. The role focuses on training, evaluating, and improving agent behavior with rigorous experimentation and measurement.

Based in San Francisco with a hybrid model (3 days in-office) and relocation support, you will collaborate across safety, capabilities, and pre-training teams to ship research-backed mitigations at scale.

Qualifications

• Strong execution and judgment in research or ML engineering projects.
• Ability to own ambiguous projects end to end.
• Interest in frontier-model safety and practical risk reduction.

Responsibilities

• Train and evaluate frontier models to reduce harmful or misaligned agent actions.
• Mine incidents and build scalable measurement, data-processing, and evaluation systems.
• Collaborate with post-training, capabilities, oversight, and pre-training teams to ship mitigations.

Skills

Research engineering

ML engineering

Quantitative research

Applied model research

Job description

OpenAI seeks a Researcher in Agent Safety, Training and Evaluations to advance safe, aligned AI agents. The role focuses on training, evaluating, and improving agent behavior with rigorous experimentation and measurement.

Based in San Francisco with a hybrid model (3 days in-office) and relocation support, you will collaborate across safety, capabilities, and pre-training teams to ship research-backed mitigations at scale.