Frontier AI Offensive Security Engineer

AI Security Institute (AISI); Greater London; Hybrid; GBP 65,000 - 145,000; Full time

Benefits offered by this job

London office

Hybrid working

Learning stipend

Conference funding

Job summary

The AI Security Institute seeks an experienced offensive security engineer to build and break across cloud, infrastructure and application layers, focusing on frontier AI model evaluations.

You will run engagements on sandboxes, Inspect and related cloud services, and you will develop agent-assisted red-teaming harnesses for continuous testing. Collaboration with engineers and researchers is expected, with opportunities to publish findings.

Qualifications

• Substantial hands-on offensive security experience in red-teaming, penetration testing, vulnerability research or exploit development.
• Strong practical ability to find and exploit vulnerabilities across cloud, container and application layers.
• Experience writing tools used by others and maintaining them.

Responsibilities

• Attack evaluation sandboxes and containment environments for frontier AI models.
• Identify, coordinate fixes, and disclose vulnerabilities in the Inspect framework and related infrastructure.
• Build and operate agent-assisted red-teaming harnesses and run them across sandbox releases.

Skills

Red-teaming

Exploit development

Python

Cloud security

Vulnerability assessment

Tools

Inspect framework

Kubernetes

AWS

Job description

The AI Security Institute seeks an experienced offensive security engineer to build and break across cloud, infrastructure and application layers, focusing on frontier AI model evaluations.

You will run engagements on sandboxes, Inspect and related cloud services, and you will develop agent-assisted red-teaming harnesses for continuous testing. Collaboration with engineers and researchers is expected, with opportunities to publish findings.