AI Red Team Engineer for LLMs (Security Certification Required)

About OpenTrain

OpenTrain is the #1 platform for people starting and growing careers in AI training and data labeling. We connect skilled contributors with hands-on projects that help build state-of-the-art AI systems and offer flexible, remote opportunities that fit around your life.

• Create a free OpenTrain account to apply in minutes and build a profile of your AI training experience.
• Work on projects that directly shape how AI systems behave while gaining practical security and ML experience.

About AI training and red teaming work

AI training (data labeling and human feedback work) is the human side of building modern models. Red teaming and security evaluation for LLMs sit at the intersection of cybersecurity and generative-AI: you simulate attacks, probe model behaviors (prompt injection, jailbreaks, etc.), and help developers harden systems.

This role is a hands-on, technical contribution to AI safety and security where your work will produce reproducible tests, automation, and clear documentation that improve model robustness.

• Contribute to cutting-edge AI safety work while working fully remote and part-time.
• Projects often require no prior annotation experience but do require relevant technical and security skills for this role.

The role

You will evaluate and red team AI models, agents, and RAG pipelines to identify safety and security vulnerabilities. This is a part-time contractor role (less than 20 hours/week) with flexible hours and remote work.

Work includes building reproducible test cases and automation, developing custom tools and test harnesses, documenting clear reproduction steps, and advising on secure coding and platform hardening.

• Part-time, remote contractor position — flexible scheduling under 20 hours/week.
• Primary focus: offensive evaluation of LLMs/agents and creation of auto-evaluable test cases and tooling.

What you’ll do

Carry out practical red team exercises and penetration tests against models, APIs, agents, and supporting infrastructure. Produce offline, reproducible tests and automation so findings are verifiable and repeatable.

Design realistic attack scenarios, create scripts and environments (Docker/CI/CD), contribute security research and write-ups, and document clear reproduction steps and mitigations for engineering teams.

• Perform web, API, network, and infrastructure penetration testing focused on AI systems.
• Develop automation, scripts and test harnesses (Python, Bash, PowerShell; Docker; CI/CD tooling).
• Create reproducible test cases and clear, actionable documentation for engineers.

Requirements and qualifications

You must have an advanced command of English (C1 or higher), and a Bachelor’s or Master’s in Computer Science, Software Engineering, Cybersecurity, Digital Forensics or a related field. Submit your resume in English.

At least one recognized cybersecurity or red team certification is required and must include a public credential.net link for verification. Applicants must reside in an eligible location (see Location & compliance).

• Degree: Bachelor’s or Master’s in CS, Software Engineering, Cybersecurity, Digital Forensics or related.
• Certification: At least one recognized security/red-team certification with a public credential.net link is required.
• Scripting: Strong skills in Python, Bash or PowerShell and experience building automation.
• Containers & CI/CD: Practical experience with Docker and CI/CD security tooling.
• Penetration testing: Hands-on testing across web, API, network and infrastructure.
• LLM risks: Familiarity with prompt injection attacks and OWASP Top 10 for LLMs; frameworks like garak or PyRIT are valuable.
• Offensive & reverse engineering: Offensive exploitation, reverse engineering tools (Ghidra) and OS-level security knowledge are valued.
• Environment: Reliable laptop and internet connection, ability to follow complex guidelines and context switch across tasks.

Preferred certifications and evidence

Preferred certifications include OffSec (OSCP, OSWE, OSEP, OSED, OSEE, OSWP), specialized red team certs (CRTP, CRTE, CRTO, CRTL, RTO, CARTP) and SANS/GIAC (GPEN, GWAPT, GXPN, GREM, GCTI, GCIH, GNFA). Foundational certs (CEH, PenTest+, CySA+, ECSA) or cloud security certs may be acceptable for more junior candidates.

Public evidence of bug bounty write-ups, CVEs, or competition results is a plus and helps demonstrate practical experience beyond certificates.

• Preferred: OffSec, CRx, SANS/GIAC family certs as listed above.
• Acceptable for junior applicants: CEH, PenTest+, CySA+, ECSA or cloud security certs (AWS/Azure/GCP specialties).
• Supporting evidence: links to public write-ups, CVEs, or competition results strengthen applications.

Location, compliance and how to apply

You must reside in an eligible location. Applications begin by confirming your current country of residence — candidates in restricted locations will not be considered. Submit your resume in English and include a public credential.net link for any required certification.

This role is offered worldwide except where restricted. Restricted countries and territories include (but are not limited to): Iran, Cuba, North Korea, Syria, Sudan, Venezuela, Myanmar; Switzerland; China, Taiwan; Kenya; Armenia, Israel, Kazakhstan, UAE, Netherlands, Serbia, Kyrgyzstan, Turkey, Uzbekistan, Belarus, Russia, Ukraine, Abkhazia, South Ossetia; many U.S. states (see job description) and a list of overseas territories. You will be asked to confirm your location during the interview process.

• Start your application by confirming your country of residence and uploading a resume in English.
• Include a public credential.net link for the required security/red-team certification.
• We will verify location and credentials during the interview process; candidates in restricted jurisdictions cannot be hired.

Compensation, schedule and engagement

This is contractor, part-time work with flexible hours and less than 20 hours per week. Compensation varies by location and experience; typical engagements reference a base hourly rate and may go up to 55 USD per hour.

Employment type: contractor/part-time. Payment details and final hourly rate will be discussed during screening and depend on experience, certifications, and location compliance.

• Time commitment: Less than 20 hours/week, flexible schedule.
• Payment: Hourly contractor pay; compensation varies by location and experience and can reach up to 55 USD/hour.
• Engagement: Short- to medium-term technical projects with potential for repeated assignments.