Frontier AI Cybersecurity Engineer – Autonomy

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

Benefits offered by this job

Hybrid working

Learning & development stipend

25 days annual leave

Pension contribution 28.97%

Opportunities to publish

Job summary

AI Security Institute in London is seeking exceptional cybersecurity experts to advance frontier AI security. You will design cyber ranges, run pre-deployment tests, and validate model behaviours, while collaborating with engineers and researchers who value autonomy and impact.

You will contribute to evaluations, reports, and publications, with mentorship from senior staff and opportunities to influence policy and practice in AI safety.

Qualifications

• Strong Python skills with automation or security tooling.
• Experience in at least one area of cybersecurity red-teaming: penetration testing, cyber range design, CTFs, automated security testing tools, bug bounties/vulnerability research/patching.
• Strong interest in helping improve the safety of AI systems.
• Familiarity with virtualization technologies such as Proxmox VE and IaC approaches is a plus.
• Ability to communicate research findings to technical and non‑technical audiences.
• Familiarity with cybersecurity tools like packet capture, testing frameworks, and reverse engineering tools.
• Active involvement in the cybersecurity community and automation/tooling impact experience.

Responsibilities

• Design cyber ranges and CTF-style challenges for evaluating AI system cybersecurity tasks.
• Build agentic scaffolding to evaluate frontier models with tools like packet capture and testing frameworks.
• Design metrics and interpret results of cyber capability evaluations.
• Collaborate with engineers to keep evaluation environments robust and scalable.
• Produce reports, papers and blogs to share findings with stakeholders.
• Keep up-to-date with related research and contribute to AI cyber risk understanding.
• Contribute to the team’s broader AI safety research initiatives.

Skills

Strong Python

Penetration testing

Cyber range design

CTFs

Automated security tools

Vulnerability research

Tools

Proxmox VE

Job description

AI Security Institute in London is seeking exceptional cybersecurity experts to advance frontier AI security. You will design cyber ranges, run pre-deployment tests, and validate model behaviours, while collaborating with engineers and researchers who value autonomy and impact.

You will contribute to evaluations, reports, and publications, with mentorship from senior staff and opportunities to influence policy and practice in AI safety.