Strategic Projects Lead - AI Safety

Leads AI safety projects for frontier AI labs, managing expert contributors and developing adversarial datasets, taxonomies, benchmarks, and red-teaming methodologies. Requires hands-on adversarial or security experience, strong writing and leadership, and production-quality Python skills.

About the job

Responsibilities

Partner directly with safety, alignment, and trust & safety teams at frontier AI labs to scope complex data problems, translate them into deliverable programs, and drive revenue.
Recruit and lead specialized contributors, including red teamers, security researchers, trust & safety practitioners, and domain risk experts.
Design specifications for adversarial datasets, jailbreak and attack taxonomies, refusal-boundary and over-refusal sets, and safety benchmarks.
Build repeatable red-teaming methodologies, including human red-team campaigns, automated attack generation, and coverage tracking against harm taxonomies.
Own projects from initial conversations through final delivery in ambiguous, fast-moving environments.
Support building, analysis, coordination, and execution as the safety practice scales.

Requirements

2+ years of experience in AI safety, red teaming, trust & safety, adversarial machine learning, or security research at a frontier AI lab, FAANG company, top security firm, or equivalent.
Hands-on adversarial experience with model jailbreaking or stress testing, prompt-injection research, offensive security, penetration testing, bug bounties, CTFs, or trust & safety investigations and enforcement.
Fluency in harm taxonomies, evaluation design, safety frameworks, and model policy.
High agency and ability to execute in ambiguity, including defining problems before solving them.
Strong writing skills and ability to specify risks precisely for others to execute against.
Genuine interest in AI and entrepreneurship.
Demonstrated track record of competitive success.
Strong leadership and communication skills.
Python proficiency and ability to write production-quality code.

Nice-to-Haves

Published safety, alignment, or security research.
Track record in bug bounties, CTFs, or model red-teaming competitions.
Experience building automated red-teaming or evaluation pipelines.
Depth in a high-consequence risk domain such as cyber, bio, fraud and financial crime, child safety, or influence operations.
Experience using AI tools to automate workflows and build products.

Compensation and Benefits

Competitive salary.
Meaningful equity.

Skills

Ai Safety, Red Teaming, Trust & Safety, Adversarial Machine Learning, Python, Prompt Injection, Penetration Testing, Bug Bounty, Ctfs, Safety Benchmarks, Evaluation Pipelines, Security Research