Staff AI/LLM Offensive Engineer (Autonomous Pentests)

Horizon3; United States; Remote; USD 235,000 - 275,000; Full time

Benefits offered by this job

Remote Work

Competitive Compensation

Equity package

Health, vision & dental insurance for你

Flexible vacation policy

Job summary

Horizon3 is hiring a Staff Attack Engineer specializing in AI/LLM security to join our remote team. You will break AI and agentic systems and translate research into automated attacks inside NodeZero, our autonomous pentesting platform.

This role focuses on building repeatable, scalable attack patterns that run autonomously across customer environments and driving our LLM-powered offensive capabilities as a technical leader.

Qualifications

• Expert Python and software engineering skills.
• Solid penetration testing fundamentals.
• Familiarity with AI/LLM security frameworks (OWASP Top 10 for LLMs, MITRE ATLAS).
• Experience in a security product or offensive security team.
• Proven ability to break AI/LLM and agentic systems.

Responsibilities

• Break AI/agentic systems and translate research into automated, repeatable attack modules for NodeZero.
• Design and execute prompt injection and defense evasion attacks with generalized, reusable patterns.
• Conduct tool-use exploitation against LLMs: access to code, files, APIs, and databases for attacker-realistic outcomes.
• Target AI infrastructure with enterprise deployment awareness and misconfigurations.
• Research model and supply chain attacks (poisoning, data extraction, adversarial inputs).
• Perform threat modeling for agentic systems and map attack surfaces to concrete paths.
• Productize manual techniques into safe, scalable automated tooling.

Skills

Python

Penetration testing

AI/LLM security frameworks

Offensive security

AI/LLM exploitation

Threat modeling

Ownership

Self-motivation

AI security trends

Tools

AWS Bedrock

AWS Agent Core

Neo4j

Azure OpenAI

GCP Vertex AI

Job description

Horizon3 is hiring a Staff Attack Engineer specializing in AI/LLM security to join our remote team. You will break AI and agentic systems and translate research into automated attacks inside NodeZero, our autonomous pentesting platform.

This role focuses on building repeatable, scalable attack patterns that run autonomously across customer environments and driving our LLM-powered offensive capabilities as a technical leader.