AI Threat Modeling & Security Researcher for Coding Agents

Apolloresearch; London; Hybrid; NOK 1,716,000 - 3,255,000; Full time

Benefits offered by this job

Flexible work hours

Unlimited vacation

Unlimited sick leave

Paid parental leave up to 6 months

Health, dental, vision insurance

Retirement matching

Lunch, dinner, snacks provided

Staff trips & conferences

Professional development budget $1,000

Relocation support and visa fees

Job summary

Apollo Research is seeking a security & control expert to design threat models for coding agents and improve the Watcher product. You will develop robust threat models, map insider-risk scenarios, and lead adversarial testing in a fast-moving, IC role that may evolve into management responsibilities.

You will work with a small team, contribute to cutting-edge AI security research, and publish analyses to communicate security posture to customers.

Qualifications

• 5+ years in security roles with hands-on technical work
• Experience in security research, threat modeling, or offensive security
• Ability to read code, review infrastructure and evaluate controls
• Strong written communication for threat models and documentation

Responsibilities

• Develop coding agent threat models including insider-risk perspectives
• Maintain and expand a library of coding agent failure modes
• Design realistic attack trajectories for monitoring and backtesting
• Serve as security adjudicator for flagged trajectories and feed insights back to product
• Lead red-teaming of Watcher monitors and contribute to product improvements

Skills

5+ years in security roles

security research

threat modeling

offensive security

written communication

Job description

Apollo Research is seeking a security & control expert to design threat models for coding agents and improve the Watcher product. You will develop robust threat models, map insider-risk scenarios, and lead adversarial testing in a fast-moving, IC role that may evolve into management responsibilities.

You will work with a small team, contribute to cutting-edge AI security research, and publish analyses to communicate security posture to customers.