SlipstreamJobs tracks this role from the company's public career site. Apply directly on the employer's site.
OpenAI's Agent Safety team is seeking a strong researcher to work on ensuring that increasingly capable AI agents act safely, exercise sound judgment, and remain aligned with user intent. The team's mission is to reduce the probability of severe unintended outcomes from frontier AI agents while preserving their ability to act effectively and autonomously.
The role spans three core areas: (1) Training—creating training methods, environments, and data that teach agents to make better decisions in consequential situations, turning real-world failures into training signals and identifying precursor behaviors and mitigations; (2) Measurements—building evaluations and production metrics that identify emerging risks and measure intervention effectiveness; and (3) Oversight—developing oversight and system mitigation mechanisms that reduce harmful actions while preserving useful autonomy.
You will train and evaluate frontier models to reduce harmful or misaligned agent actions, forming clear hypotheses and executing independently through ambiguity. You'll mine incidents and build scalable measurement, data-processing, and evaluation systems that turn real failures into repeatable safety signals. Close collaboration with post-training, capabilities, oversight, and pre-training partners is essential to ship research-backed mitigations into large-scale training and agent systems.
Ideal candidates have demonstrated strength in research engineering, ML engineering, quantitative research, or applied model research, with the ability to own ambiguous projects end-to-end. Strong technical execution across experimentation, data, evaluation, and/or infrastructure is required, along with strong intuition for modern frontier-model research. Prior safety or alignment experience is not necessary; the team welcomes people who have recently realized that alignment and safety is a critical area to contribute to.
This is a hybrid role based in San Francisco, requiring 3 days in the office per week. OpenAI offers relocation assistance to new employees.
