Frontier AI Safety Researcher: Mitigations & Alignment

OpenAI; San Francisco (CA); On-site; USD 180,000 - 280,000; Full time

Job summary

OpenAI in San Francisco is seeking exceptional researchers to push the frontier of safety mitigations for deployed models. You will help derisk frontier models by developing novel safety mitigations and applying methods from interpretability, control, and alignment to ensure safe AI systems.

The role requires strong technical depth and collaboration across misalignment, cybersecurity, and biology to build an end-to-end safety stack that is scalable and enforceable at scale.

Qualifications

• 2+ years of experience in AI safety research or related field.
• PhD or equivalent in CS/ML is required.
• Strong programming in Python and familiarity with ML libraries.

Responsibilities

• Identify emerging AI safety risks and mitigation approaches.
• Develop and refine evaluations to measure risk and safety.
• Set research directions to improve AI safety, alignment, and robustness.
• Contribute to best practices for AI safety across OpenAI and industry.
• Design red-teaming pipelines to test end-to-end safety systems.

Skills

AI safety

RLHF

Interpretability

Robustness

Control

Python

Education

PhD or equivalent in CS/ML

Tools

PyTorch

TensorFlow

SciPy

Job description

OpenAI in San Francisco is seeking exceptional researchers to push the frontier of safety mitigations for deployed models. You will help derisk frontier models by developing novel safety mitigations and applying methods from interpretability, control, and alignment to ensure safe AI systems.

The role requires strong technical depth and collaboration across misalignment, cybersecurity, and biology to build an end-to-end safety stack that is scalable and enforceable at scale.