Researcher, Frontier Cybersecurity Risks

Highlights

• Build end-to-end mitigation stack for severe model-enabled cyber misuse.
• Stress-test safeguards against novel exploits and automated attack workflows.
• Help keep frontier AI safeguards ahead of evolving attacker behaviors.

Ideal candidate

Safety-motivated researcher who loves building robust protections for frontier AI and thrives collaborating across research, product, and security to reduce severe cyber misuse.

Requirements

• Passion for AI safety and safer real-world use of cutting-edge AI models
• Demonstrated experience in deep learning and transformer models
• Proficiency with frameworks such as PyTorch or TensorFlow
• Strong foundation in data structures, algorithms, and software engineering principles
• Familiarity with training and fine-tuning LLMs including distillation, supervised fine-tuning, and policy optimization
• Significant experience designing and deploying technical safeguards for abuse prevention, detection, and enforcement at scale
• Ability to collaborate effectively across research, security, policy, product, and engineering
• Background knowledge in cybersecurity or adjacent fields (nice to have)