Strategic Projects Lead, Safety
Core
Own end-to-end execution of large-scale adversarial testing and red-teaming programs for frontier AI labs to push models to failure and develop safety policies.
Role type
Senior IC strategic projects lead (AI safety & red teaming)
Builds
Adversarial testing pipelines, jailbreak techniques, agentic red-teaming methods, and safety policy frameworks for frontier AI models
Domain
AI Safety, Red Teaming, Frontier AI Model Evaluation
Deliverable
production ML models
Required skills
Project and workforce management, analytical problem-solving, adversarial testing concepts, stakeholder management, high ownership mindset, rapid learning of technical AI concepts
Preferred skills
LLM red-teaming and jailbreaking experience, expertise in high-severity policy harms (CBRN, cyber, violent extremism), running testing/evaluation programs, security research background
Technologies
LLMs, agentic AI systems, adversarial testing frameworks
Responsibilities
Design and run adversarial testing methodologies (jailbreak, push-to-failure), extend red-teaming to agentic contexts, lead teams of expert Fellows, partner with policy leads to shape safety documents, synthesize testing data into insight reports, design staffing models for red-teaming workforce
Seniority
Senior, hands-on IC