About the role

Cobalt builds expert data and evaluation infrastructure for AI developers. We are recruiting senior cybersecurity practitioners for a contract project that tests the limits of frontier AI models on real security work. You will act as an adversary to the model: your job is to find the places where it breaks by designing realistic, hands-on security challenges in a command line environment that a leading AI model cannot solve. Accepted challenges are used to evaluate and train the next generation of frontier models.

What you will do

Design self-contained command line challenges based on the work you have done in your career, such as penetration testing, exploit development, malware analysis, reverse engineering, network forensics, incident response, and hardening of misconfigured systems.
Build the challenge environment, write a reference solution, and write automated tests that verify whether a solution is correct.
Run your challenge against a frontier AI model, study how it fails, and refine the challenge until the failure reflects a real gap in the model's capability rather than ambiguity or trick wording.
Work with reviewers to bring each challenge to acceptance.

Who we are looking for

A PhD in computer science, cybersecurity, or a closely related field.
Industry or academic experience in offensive security, defensive security, or security research.
At least one publication, either academic (for example a peer-reviewed paper) or professional (for example a conference talk, a published vulnerability disclosure, or a widely used open-source tool).
Fluency in the Linux command line, shell scripting, Docker, and Python.
Experience designing capture the flag challenges, training labs, or penetration test scenarios is a strong advantage, because the work is very similar.
The ability to write precise specifications that another expert could follow without asking questions.

Why Cobalt AI:

Advance frontier AI where it counts. Apply your research expertise to the data that frontier labs cannot obtain any other way, where your reasoning directly shapes how the next generation of models works through technical problems.
Grow professionally.

Expand your influence through evaluation projects, advisory roles, and research collaborations, while deepening your understanding of how frontier models are trained and assessed.
Work with a top-tier network.

Collaborate with researchers from leading institutions and labs on high-impact, flexible work.
Set your own schedule.

Flexible 10 to 40 hour weeks that fit around your research position and your life.
Competitive pay.

Rates vary by project and are determined by a number of factors, including scope, skillset, and experience.

Senior Cybersecurity Expert, AI Training and Red Teaming (PhD, Remote)

Job description