AI Tester

Information Security Specialist – AI Penetration Tester | Remote Role Summary The Information Security Specialist – AI Penetration Tester performs hands-on adversarial security testing of AI/ML and GenAI solutions, including LLM applications, RAG architectures, AI agents, model pipelines, APIs, and supporting cloud infrastructure. The role identifies and validates AI-specific vulnerabilities, assesses the effectiveness of security controls and guardrails, and works with technology teams to remediate identified risks. This lines up directly with the operating model you subsequently described for the position: AI Penetration Testers / Red Teamers conduct required manual security testing of AI systems, platforms and agents using adversarial techniques to identify vulnerabilities, validate controls and support remediation. Key Accountabilities AI Penetration & Adversarial Testing: Design and execute security testing against AI/ML models, LLM applications, RAG systems, AI agents, model pipelines and AI-enabled business workflows. AI Vulnerability Identification: Identify vulnerabilities including prompt injection, jailbreaking, model extraction, adversarial ML attacks, data poisoning, RAG bypasses and safety/guardrail circumvention. Agentic AI Security Testing: Assess agent autonomy and AI-enabled systems, including their security controls and complete AI attack paths. Your internal requirements also specifically call for experience with AI agentic penetration testing and kill-chain development. Traditional Penetration Testing: Apply strong application-security and penetration-testing fundamentals alongside AI-specific techniques. Your internal hiring guidance identifies AppSec penetration testing and traditional red/purple-team experience as important foundations for this role. Tooling & Automation: Evaluate and develop internal and open-source tooling to automate and scale AI/LLM security testing, with Python identified as a preferred scripting capability. AI Security Test Harnesses: Use security testing harnesses and AI-assisted offensive-security capabilities to improve testing depth and attack coverage. Risk Assessment & Hardening: Assess inference endpoints, system prompts, guardrail designs, agent autonomy, model monitoring and ModelOps pipelines. Model Supply Chain Security: Evaluate AI supply-chain risks, dependency integrity and alignment with enterprise security requirements. Findings & Remediation: Document vulnerabilities, communicate technical and business impact, and work with Engineering, Data Science, Cloud, Cyber Defense, Architecture and Risk stakeholders to remediate findings. Standards & Governance: Contribute to AI security standards, testing methodologies, control objectives and hardening practices. Required / Preferred Experience The existing JD calls for: 5+ years in application security, penetration testing or red teaming, with hands-on AI/ML security experience preferred. Knowledge of AI/LLM vulnerabilities, OWASP guidance, adversarial attacks, prompt injection and model-safety testing. Familiarity with Python, model interrogation techniques, and cloud AI environments across Azure, AWS or GCP. Familiarity with frameworks including NIST AI RMF, OWASP LLM/ML guidance, ISO 42001 and MITRE ATLAS. Relevant security certifications such as OSCP, GPEN, COASP, GAIPT with AI/ML security certifications also considered relevant.