AI Security Researcher: Insider Threat & Threat Modeling

Apolloresearch; London; Hybrid; NOK 1,822,000 - 2,391,000; Full time

Benefits offered by this job

Health, dental and vision insurance

Unlimited vacation

Unlimited sick leave

Parental leave up to 6 months

Meals provided

Retirement matching (401k)

Relocation and visa support

Professional development budget $1,000

Job summary

Apollo Research is hiring AI Security Researchers to join the Infra & Security Team in London or San Francisco. You will identify, research and remediate risks introduced by AI agents that could affect Apollo and our mission, redefine insider risk, and build durable security controls.

You’ll red-team internal software and AI agent access controls, track adversary techniques, and own findings through deployment, collaborating with engineers to raise the security bar in a fast-paced startup

Qualifications

• 5+ years in security roles in a hands-on technical capacity.
• Direct experience with offensive security, threat modelling, red teaming, and cloud or application security.
• Engineering mindset: translate findings into fixes, automation, CI/CD, and scalable controls.
• Startup pace: able to work in a fast-moving environment with ambiguity.
• Strong written communication: produce threat models and reports clearly.

Responsibilities

• Red-team internal software, infrastructure, and AI agent access controls/monitoring.
• Design solutions for novel or emerging threats in the AI security space.
• Track adversary tactics and translate intelligence into curated detections.
• Own findings through to deployed fixes and socialize them with engineers.

Skills

5+ years in security roles

Offensive security

Engineering mindset

Startup pace

Strong written communication

Job description

Apollo Research is hiring AI Security Researchers to join the Infra & Security Team in London or San Francisco. You will identify, research and remediate risks introduced by AI agents that could affect Apollo and our mission, redefine insider risk, and build durable security controls.

You’ll red-team internal software and AI agent access controls, track adversary techniques, and own findings through deployment, collaborating with engineers to raise the security bar in a fast-paced startup