AI Security Red Teamer

Job summary

Design and conduct authorized security tests against AI models, agents, applications, and infrastructure, investigating vulnerabilities such as prompt injection, jailbreaks, unintended tool use, and data exposure. Create reproducible test cases, assess findings, recommend mitigations, collaborate on retesting, and document results clearly.

Job details

Help leading AI labs build safer, more resilient models. Parallel connects experienced cybersecurity specialists with projects supporting leading AI labs. We’re hiring seven AI Security Red Teamers in United States to uncover vulnerabilities in frontier AI systems and produce findings that help improve model training, evaluation, and security. You’ll apply offensive security expertise to large language models, image generation systems, AI agents, and their supporting infrastructure. Your work will turn successful attacks and unexpected model behavior into reproducible test cases, actionable reports, and recommendations for stronger defenses. This is remote, project-based work. Engagements and hours vary with project demand, with scope and expected availability discussed before each assignment. What you’ll do Design and execute authorized adversarial tests against generative AI models, agents, applications, and infrastructure. Investigate prompt injection, jailbreaks, unintended tool use, data exposure, and other AI security vulnerabilities. Apply web penetration testing techniques to the applications and services surrounding AI systems. Build reproducible attack scenarios and evaluation cases that support model training and security testing. Assess the impact of findings and recommend practical mitigations. Collaborate with project teams to validate weaknesses and retest improvements. Document your methods, evidence, and results clearly enough for another practitioner to reproduce. What you’ll bring At least five years of web penetration testing experience. Demonstrated experience identifying and analyzing AI vulnerabilities. Strong understanding of generative AI models, agentic frameworks, and their underlying architectures. Strong analytical judgment, creative problem-solving, and clear written communication. Ability to work independently, adapt to changing project requirements, and communicate reliably with a remote team. Ability to pass Parallel’s practical CTF assessment. Preferred experience Python or Node.js for scripting, automation, and security tooling. Building or testing AI agents and agentic workflows. Advanced offensive-security certifications such as OSWE or OSCE3. Published CVEs, responsible disclosures, bug bounty findings, or strong CTF results. A relevant degree in computer science, information security, or a related field. Selection process Applicants must pass a practical CTF assessment and complete a background check before engagement. Final compensation within the posted range will reflect relevant experience and demonstrated technical ability.

What you’ll do

Design and conduct authorized security tests against AI models, agents, applications, and infrastructure, investigating vulnerabilities such as prompt injection, jailbreaks, unintended tool use, and data exposure. Create reproducible test cases, assess findings, recommend mitigations, collaborate on retesting, and document results clearly.

Requirements

Applicants need at least five years of web penetration testing experience, demonstrated experience analyzing AI vulnerabilities, and a strong understanding of generative AI models, agent frameworks, and their architectures. They must communicate clearly, work independently with a remote team, and pass a practical CTF assessment; Python or Node.js experience and relevant certifications or disclosures are preferred.

Listed skills

• Python · Preferred
• Node.js · Preferred
• Technical Documentation · Preferred

Other relevant skills

Identified from the job description. Confirm important requirements above.

• Web Penetration Testing
• AI Vulnerability Analysis
• Generative AI Security
• Adversarial Testing
• Prompt Injection Testing
• Jailbreak Testing
• AI Agent Security
• Application Security
• Attack Scenario Development
• Security Evaluation
• Risk Assessment
• Mitigation Recommendations
• Python
• Node.js
• Technical Documentation

Job areas

• Security & Safety
• Technology
• Software

More jobs from Parallel