Researcher, Multimodal Safety

The role

You will define and advance multimodal safety research for text, vision, and audio. You will build training and evaluation methods for VLMs and collaborate with various teams to translate research into safer multimodal experiences.

What you will do

• Define and advance multimodal safety research for text, vision, and audio.
• Build training and evaluation methods for VLMs, including post-training, safety evals, and interventions.
• Collaborate with Personal AGI, Consumer Devices, and product/model teams to translate research into safer experiences.

Required qualifications

• Track record of building or advancing multimodal models, with depth in vision-language models, video understanding, image generation, audio, or multimodal reasoning.
• Fluency across both perception and language.
• Understand how multimodal systems work end to end, from encoders, projection layers, and modality fusion to cross-modal reasoning, scaling, and inference tradeoffs.
• Improved frontier model behavior through post-training, using approaches such as SFT, RL, data curation, synthetic data, evaluation, and rigorous error analysis.
• Strong research and engineering judgment to form testable hypotheses, design experiments, diagnose model failures, and translate findings into improvements.

Benefits

• Relocation assistance to new employees

Before you apply

• Background checks for applicants will be administered in accordance with applicable law.

About the company

OpenAI is an AI research and deployment company dedicated to ensuring that general-purpose artificial intelligence benefits all of humanity. They seek to safely deploy AI through their products.