Researcher, Health AI

JOB DESCRIPTION

OpenAI is seeking a Researcher to join its Health AI team within the Safety Systems organization. This role focuses on advancing the safety, robustness, and reliability of AI models designed for healthcare and medical information use cases. The team operates at the intersection of AI safety research and real-world health applications, with the goal of enabling universal access to trustworthy, high-quality medical information.

In this role, you will conduct applied and foundational research to improve the behavior, reasoning, and safety properties of large language models used in health-related settings. You will design and evaluate scalable alignment and safety techniques, ensuring models provide accurate, reliable, and responsible outputs when deployed in sensitive medical contexts.

You will work closely with researchers, engineers, and cross-functional partners to integrate safety improvements into core model training pipelines and product launches. This includes developing reusable tools and libraries for alignment methods, proactively identifying risk areas, and contributing to safety strategies that generalize toward safe and beneficial AGI.

The position requires strong ownership, the ability to work end-to-end on complex research problems, and a collaborative mindset. Candidates with experience in deep learning, large language models, and practical alignment techniques such as reinforcement learning from human feedback, automated red teaming, or scalable oversight will thrive in this environment. Experience in health-related AI research or deployments is considered a strong plus.

This role is based in San Francisco and follows a hybrid work model, with relocation assistance available.