Researcher, Automated Red Teaming

Responsibilities

• Lead the Automated Red Teaming (ART) effort by building scalable, research-driven systems to uncover model failure modes
• Own research and technical direction for automated red teaming in catastrophic risk areas like cyber and bio threats
• Partner with vertical risk teams and classifiers teams to turn discovered attacks into training data and measurable robustness gains

Requirements

• Hands-on experience with LLMs and agents, including multi-turn behaviors and tool use
• Strong applied research instincts, particularly in designing reproducible and interpretable evaluations
• Solid software engineering fundamentals including data structures, algorithms, and testing discipline

Preferred Qualifications

• Experience in adversarial ML or security research
• Experience with red teaming or abuse prevention systems
• Experience building large-scale evaluation infrastructure

About the Company

OpenAI is an AI research and deployment company dedicated to ensuring that general-purpose artificial intelligence benefits all of humanity.