AI Security Analyst – III

Overview

This position will be responsible to lead implementation and continuous improvement of our AI Security Program. You will design controls, define detections, and drive remediation for AI agents/applications spanning Purview DSPM for AI, Wiz DSPM for AI, Azure AI Foundry, Copilot Studio, Microsoft AI Gateway, Agent 365, Entra ID, and Azure networking—with centralized governance, auditing, and incident response in Sentinel/Defender/Security Copilot

Key Responsibilities

Own the AI security control catalog and lifecycle controls, mapping them to Microsoft/Azure capabilities and enterprise policies.

Define standards for non‑human identity (Agent ID/Managed Identity), Conditional Access, and least‑privilege patterns for agents/services; partner with platform teams to automate provisioning.

Set network isolation patterns for AI endpoints (Private Link, deny public endpoints, controlled egress).

Establish release gates that require Foundry safety filters/evaluations and Purview policy checks for high‑risk AI use cases.

Design and tune AI‑specific detections (prompt injection, jailbreak attempts, sensitive‑data exfil) in Sentinel/Defender.

Lead incident investigations for AI apps/agents, coordinate cross‑functional response, and deliver executive‑level narratives and root‑cause analyses.

Drive DSPM strategy across Purview and Wiz (classification, labeling/auto‑labeling, access/usage policies, posture scoring); reduce exposure in both SaaS and internal environments.

Work collaboratively with IT and business stakeholders to align Cybersecurity initiatives with broader organizational goals and objectives.

Provide technical input and support in resource planning and roadmap execution for Cybersecurity and AI governance initiatives.

Requirements

Strong proficiency with Purview (DLP/Information Protection/Insider Risk), Wiz DSPM for AI, Sentinel, Defender, Entra ID/Conditional Access, APIM/Gateway, Azure networking (VNets/Private Link).

Experience integrating Azure AI Foundry and Copilot Studio with security guardrails and release gates; understanding of Agent 365 concepts and agent governance.

Familiarity with Cranium (AI risk assessments) and Cyera/Reco for data discovery and classification at scale.

Threat modeling for AI/agentic apps; crafting content‑safety/evaluation policies and red‑team scenarios.

Strong knowledge of AI technologies, cybersecurity principles, and risk management practices.

Excellent communication and interpersonal skills, with the ability to engage and influence stakeholders across various levels of the organization.

Foundational skills in data engineering, cybersecurity, intelligent automation, machine learning and emerging technologies.

Ability to analyze complex situations, propose solutions, and drive continuous improvement initiatives.

Qualification

Bachelor’s degree or higher in Computer Science, Cybersecurity, Engineering or a related field.

5-7+ years in Cloud/Data/App Security with 3+ years hands on in M365/Azure security; experience leading security control design/implementation.