Frontier AI Safety Researcher

Triwill Group; San Francisco, Northern (CA, KY); Hybrid; USD 190,000 - 230,000; Full time

Benefits offered by this job

Relocation assistance

Job summary

OpenAI is seeking strong executors with excellent judgment for a role focused on training and evaluating frontier models to reduce harmful agent actions. The position emphasizes building scalable measurement and evaluation systems and collaborating across post-training, capabilities, oversight, and pre-training teams.

Hybrid work in San Francisco, CA with relocation assistance, welcoming candidates even if they come from outside formal safety or alignment backgrounds.

Qualifications

• Strong executors with excellent judgment and comfort with ambiguity.
• Understanding of frontier model research.
• Open to contributing even without prior safety/alignment experience.

Responsibilities

• Train and evaluate frontier models to reduce harmful or misaligned agent actions, forming hypotheses and executing through ambiguity.
• Mine incidents and build scalable measurement, data-processing, and evaluation systems for safety signals.
• Collaborate with post-training, capabilities, oversight, and pre-training teams to ship mitigations into large-scale training and agent systems.

Skills

Research engineering

ML engineering

Quantitative research

Applied model research

Experimentation

Data processing

Evaluation

Infrastructure

Ambiguity tolerance

Safety/alignment interest

Tools

Post-training coordination tools

Large-scale training systems

Job description

OpenAI is seeking strong executors with excellent judgment for a role focused on training and evaluating frontier models to reduce harmful agent actions. The position emphasizes building scalable measurement and evaluation systems and collaborating across post-training, capabilities, oversight, and pre-training teams.

Hybrid work in San Francisco, CA with relocation assistance, welcoming candidates even if they come from outside formal safety or alignment backgrounds.