jobs

Researcher, Safety Training, National Security

About the TeamThe Safety Training research team aims to fundamentally advance our capabilities for precisely implementing safe behavior in AI models, and to leverage these advances to make OpenAI’s deployed models safe and beneficial. This

Company

OpenAI

Location

San Francisco; Washington, DC; US - Remote

Status

Open

Posted

2026-09-12T03:21:15.873+00:00

OpenAI's Safety Training research team is hiring a researcher to train and evaluate models for U.S. government use, focused on national security applications. The work covers safety training methods, reinforcement learning and adversarial robustness, building evaluations, identifying failure modes and feeding findings back into training, alongside research, engineering, security and policy partners. The posting asks for 4+ years of AI safety research experience including RLHF or adversarial training, a degree in computer science, machine learning or a related field, strong deep learning skills, and an active TS/SCI clearance.