Principal AI Security Analyst

Job description

Overview As a Principal AI Security Analyst, you will be the organization's deepest subject-matter expert at the intersection of artificial intelligence and cybersecurity. You will lead the security review of AI/ML systems, manage our AI security toolchain, identify emerging threats unique to large-scale models, and build the frameworks that keep our AI infrastructure resilient against adversarial attacks. This is a high-visibility, high-autonomy role that shapes how the company thinks about AI risk. Responsibilities Own end-to-end threat modeling for AI and ML systems, including LLMs, training pipelines, and inference infrastructure Administer and tune Palo Alto Networks AI Runtime Security (AIRS) to detect and block adversarial inputs, prompt injection, and model abuse in real time Manage AI Access Security policies to govern employee use of third-party AI applications - enforcing DLP, acceptable-use rules, and shadow-AI visibility Integrate an AI gateway layer to apply rate limiting, access controls, and observability across LLM API traffic Research and operationalize defenses against adversarial attacks - model extraction, data poisoning, jailbreaking, and membership inference Lead red-team exercises targeting AI systems and synthesize findings into actionable security roadmaps Define and maintain security standards, policies, and controls specific to AI model development and deployment Partner with ML engineering, platform, and product teams to embed security requirements from design through production Evaluate third-party AI tools, APIs, and vendors for supply-chain and data-handling risk Work with other security team members in AI governance discussions, including compliance with EU AI Act and NIST AI RMF Mentor other team members; set technical direction for the AI security practice Qualifications REQUIRED 8+ years in cybersecurity with 3+ years focused on AI/ML security Hands-on experience with Palo Alto Networks AIRS or AI Access Security Deep understanding of LLM architectures and common vulnerability classes Proficiency in Python; ability to review model code and ML pipelines Experience with threat modeling frameworks (STRIDE, PASTA, or similar) Track record driving cross-functional security programs at scale Cloud-native environment experience (AWS, GCP, or Azure) PREFERRED Palo Alto Networks certifications (PCNSE, PCCSE, or equivalent) Experie…