Back to Careers
Careers

Senior Security Researcher

Offensive AI Security Research · SpartanX Labs

Remote (Americas preferred)Full-timeReports to: Chief Research Officer (CRO)

About the role

SpartanX Labs finds tomorrow's attacks before they become today's incidents. We attack AI systems, measure how well autonomous attackers perform, and push what we learn into the product. This is not a pentest job. You will lead research, publish original work, and help build one of the first dedicated offensive AI research teams in the industry, working directly with our CRO.

What you will do

Break things

  • • Find and develop new attack techniques against AI systems, cloud, applications, and enterprise networks.
  • • Build proof-of-concept exploits and attack chains.
  • • Research prompt injection, agent abuse, and model exploitation.
  • • Test our offensive AI agents against realistic attack scenarios.
  • • Track real attacker tradecraft and turn it into new tests.
  • • Run responsible disclosure and report vulnerabilities (CVEs) when it applies.

Measure things

  • • Build and maintain repeatable benchmark environments to score our AI attacker.
  • • Define success criteria, scoring, and metrics. Re-run them to show progress over time.
  • • Wire telemetry and automation into dashboards so results are visible and trusted.
  • • Build the tooling and automation that makes research repeatable.

Ship research into the product

  • • Work with product and engineering to turn findings into product improvements.
  • • Use research to inform roadmap decisions.
  • • Publish original research: blogs, papers, conference talks, and CVEs.
  • • Mentor other researchers and raise the bar on research quality.

Your first 90 days

  • • Day 30: Own and lead one research workstream and ship a first measurable result.
  • • Day 90: Lead that workstream end to end and mentor others on it.

Workstreams you could own:

  • • Benchmark Lab: vulnerable lab targets we use to score our AI attacker.
  • • Offensive Agent Evaluation: testing how well our AI agents attack, and where they fail.
  • • Novel Attack Research: finding new attack techniques, especially against AI systems.
  • • Bug Bounty: running our program and finding real bugs in real targets.

Must-haves

  • • 5+ years in offensive security, red teaming, or vulnerability research.
  • • Strong web, cloud, operating system, and network security fundamentals.
  • • Scripting in Python, PowerShell, Bash, or similar.
  • • Hands-on with Burp Suite, Nmap, Metasploit, and common offensive tooling.
  • • A track record of original research: published blogs, talks, CVEs, or similar.
  • • You research problems on your own and explain findings clearly.
  • • You use AI tools to move faster, and you validate their output and handle sensitive data properly.

Nice to have (one or more)

  • • AI security, agentic AI, prompt injection.
  • • Cloud security (AWS, Azure, GCP) and Kubernetes.
  • • Active Directory, threat research, malware analysis, detection engineering.
  • • Benchmark development, infrastructure as code, security automation, bug bounty.

Certifications (not required): OSCP, PNPT, HTB CPTS, CRTO, eCPPT, AWS Security Specialty.

What we look for

  • • Curiosity. You want to know how things break.
  • • Deep offensive fundamentals.
  • • A research mindset and attention to detail.
  • • Clear writing and speaking.
  • • You work well alone and with a team.
  • • A bias to experiment, measure, and improve.
  • • You use AI to go faster without losing rigor or accuracy.

Career path

LevelFocus
Senior Security ResearcherLead major research workstreams, publish original work, mentor researchers.
Principal Security ResearcherSet technical direction and represent SpartanX externally.

What we offer

  • • Work at the front line of offensive AI.
  • • Help build one of the first dedicated AI security research teams.
  • • Hard, real problems: attack chains, evaluation, reproducibility, tradecraft.
  • • High ownership, radical transparency, zero drama.
  • • Flexibility. We judge results, not hours.

About SpartanX

SpartanX is the first Autonomous Exposure Management (AEM) platform: a continuous adversary that proves the ways into your whole attack surface, inside and out, and helps you fix them. Exploit-validated, always on, and under your control.