AI Security & Information Protection Analyst – Agentic AI

Key Responsibilities:

• Assess agentic AI use cases, including LLM agents, tool/function calling, orchestration, and RAG, for prompt injection, data exfiltration, model abuse, unsafe tool invocation, and supply-chain risks. 

• Support AI security guardrails, secure configurations, role/data scoping, allow/deny policies, reference architectures, and usage guidance. 

• Administer and tune DLP controls across endpoints, email, collaboration platforms, cloud, and SaaS; monitor, triage, escalate, and remediate alerts with data owners and legal/privacy teams.

• Improve data classification and labelling and detect unsanctioned AI apps, risky plugins, browser extensions, shadow AI, and data uploads. 

• Support investigations using agent logs, tool-call traces, and prompt/response histories; develop playbooks and AI security KPIs/KRIs. 

• Contribute to AI policies, training, third-party due diligence, and vendor/SaaS assessments covering residency, retention, training-data controls, auditability, and transparency. 

Required Qualifications: 

• Foundational knowledge of agentic AI, RAG, prompt injection, leakage pathways, DLP policies/classifiers/incident workflows, IAM, least privilege, network and SaaS security. 

• Experience with at least one enterprise DLP/CASB/SSE platform, SIEM/EDR/cloud-security monitoring, or SaaS governance is required. 

• Strong analytical, documentation, stakeholder communication, adaptability, and learning mindset are essential. 

Desirable Skills: 

• Exposure to LLM ecosystems, agent frameworks, vector stores, model gateways, prompt-injection and exfiltration patterns, DLP classifier tuning, EDM/IDM fingerprints, privacy requirements, SaaS risk assessments, Python/PowerShell, logs/JSON, and NIST AI RMF, ISO/IEC 23894, OWASP LLM guidance, or MITRE ATLAS. 

• Success in the First Six Months Baseline sanctioned and unsanctioned AI usage; improve AI-focused DLP precision; publish secure AI guidance; implement initial detections/playbooks for prompt Information Protection Analyst (Agentic AI Focus) – 6Month OSP Engagement injection, risky tools, and abnormal data movement; and complete at least one AI vendor or solution risk assessment. 

• Engagement Details OSP contractor; six months, with possible extension; hybrid/remote per terms. 

• Limited on-call or incident-support windows may apply. 

• Candidates should be comfortable with ambiguity, practical incremental improvement, and learning by doing while protecting sensitive data and enabling business value.