Researcher, Automated Red Teaming
OpenAI
3 days ago
San Francisco, CA, USA
Mid Level / Senior
Base Salary
$295k - $445k/yr
Responsibilities
- Lead the Automated Red Teaming effort to discover failure modes in AI models.
- Build scalable, research-driven systems for continuous risk assessment.
- Collaborate with vertical risk teams to define threat models and prioritize targets.
- Work with the Classifiers team to convert discovered attacks into training data.
- Ensure ART outputs are operationally useful for product and engineering stakeholders.
- Translate findings into actionable improvements and communicate effectively with various teams.
Requirements
- Strong motivation towards AI safety and reducing catastrophic risks.
- Experience in breaking systems and identifying high-severity failure modes.
- Hands-on experience with LLMs and agents, including multi-turn behaviors.
- Ability to build scalable automation and turn ideas into continuous pipelines.
- Solid software engineering fundamentals, including data structures and algorithms.
- Experience in adversarial ML, security research, or large-scale evaluation infrastructure is a plus.
Categories
AI & MLSecurityTesting