Horizon3AI
Staff Attack Engineer, AI/LLM
$235K–$275K
US, Remote · Remote · Staff · Full-time · Posted Apr 10, 2026
About the job
Break AI and agentic systems and turn the research into scalable automated attack modules for NodeZero, while building LLM-powered offensive capabilities.
What you'll do
- Research and automate attacks against AI/LLM systems, agents, and AI infrastructure.
- Build LLM-powered applications and microservices with production concerns such as reliability, safety, cost, and observability.
- Provide technical leadership for AI/LLM offense and drive high-complexity programs.
What they ask for
- Expert-level Python and software engineering skills.
- Solid penetration-testing fundamentals and understanding of common attack chains.
- Familiarity with AI/LLM security frameworks, such as OWASP Top 10 for LLMs or MITRE ATLAS.
- Proven ability to break AI/LLM and agentic systems.
Nice to have
- Experience with cloud AI services, AWS Bedrock, AWS Agent Core, or graph databases such as Neo4j.
- AI security research contributions, exploit development, vulnerability research, or AI/ML-focused CTF experience.
Pay
$235,000 - $275,000 annually (base salary)
Tools
Python, OWASP Top 10 for LLMs, MITRE ATLAS, Azure OpenAI, GCP Vertex AI, AWS Bedrock, AWS Agent Core, Neo4j
AI focus
Offensive AI/LLM security research and automated attacks, including prompt injection, tool-use exploitation, AI infrastructure and supply-chain attacks, and LLM-powered applications and workflows.
Model-written summary · the original posting is authoritative