Principal AI Security Engineer

CDW

21h•$172,000 - $240,000

About The Position

Join CDW and help protect delivery of full stack technology solutions and global services for 250K+ customers—including corporate enterprise, government, education, and healthcare industries. You will be on a team dedicated to collaborative delivery of a new global information security strategy, operating model, and objectives to accelerate CDW’s business goals in a secure way. Your role at CDW is of the utmost importance to the company’s mission, objectives, and reputation. As the Principal AI Security Engineer, you will be the technical authority on keeping AI secure across the enterprise. At the heart of this role is assessment: you'll own how CDW evaluates AI risk before it reaches production, then design and build the controls that close it. You’ll partner across our security teams and platform teams to define the guardrails, identity model, and detection capabilities that let CDW use AI aggressively and securely, whether we build it, buy it, or need to block it. This is a hands-on principal role. You will set enterprise AI security direction and help build what you design: reference architectures, controls, and tooling. You will also serve as a trusted advisor across CDW, helping senior leaders and AI governance teams make practical, risk-informed decisions as we scale AI adoption. Your impact: fewer AI-related incidents, faster and safer AI adoption, and security controls that scale with the business.

Requirements

Bachelors degree and 10+ years in security engineering or security architecture, including hands-on work securing AI/ML systems, or 14+ years of experience in security engineering or security architecture, including hands-on work securing AI/ML systems.
Proven experience designing and leading security assessments of AI systems, platforms, or vendors, and turning findings into clear, risk-based decisions.
Hands-on experience securing production AI technologies, including at least one of the following: Deploying AI guardrails or sandboxes, Managing AI or agent identities, Detecting or mitigating prompt injection, Implementing AI security tooling, built or bought
Experience building or evaluating security for production AI systems, beyond research or theory.
Solid understanding of how LLMs work, including pretraining, fine-tuning, RLHF, inference, retrieval-augmented generation, and safety alignment, and how each layer can be attacked.
Experience securing AI platforms and agent frameworks such as Microsoft Copilot Studio, Azure AI services, OpenAI or Anthropic tool use, MCP-based integrations, and frameworks such as Semantic Kernel or LangChain.
Experience securing AI deployments in public cloud (Azure, AWS, or GCP), identity and access management, and zero trust principles.
Proven ability to influence senior leaders and cross-functional teams without direct authority.

Nice To Haves

Experience on a security team defending production AI systems at scale, or at a frontier AI lab, a plus.
Background in Infrastructure, Cloud, Platform, or Corporate Security, Security Operations, or Detection and Response, a plus.
Software security review, secure SDLC, vulnerability discovery, or offensive security experience, a plus
Published work, conference talks, CTF results, or production red team work in prompt injection, jailbreak research, or adversarial ML, a plus.
Experience with AI security posture management, AI gateways, or AI observability platforms, a plus.
Relevant certifications such as CISSP, CCSP, OSCP, or AI security credentials, a plus.

Responsibilities

Own CDW's AI security assessment program. Design and mature the process, criteria, and risk tiers used to approve, restrict, monitor, or block AI use cases, platforms, vendors, and shadow AI.
Lead security assessments of AI systems, agents, and integrations before and after deployment, and set the security bar other teams build to.
Partner with the AI Council, AppSec, and platform teams to build assessment findings into AI use case reviews, so risks are found and addressed before deployment.
Evaluate third-party AI vendors, purchased AI capabilities, and unapproved AI use against CDW's AI security requirements and guardrails.
Define methods and tools for AI red teaming, including automated adversarial testing, to validate assessment findings and test controls under real attack conditions.
Assess AI-related controls across CDW's existing security stack, find gaps, and drive root-cause remediation rather than one-off fixes.
Define and own CDW's enterprise AI security architecture, including standards, reference architectures, and secure design patterns aligned to NIST AI RMF, ISO/IEC 42001, OWASP Top 10 for LLM Applications, and MITRE ATLAS.
Set security requirements across the AI lifecycle, from data handling and model selection through deployment, monitoring, and retirement.
Define trust boundaries, authorization models, and least privilege for single-agent and multi-agent systems.
Design end-to-end security for the AI stack, including LLM inference pipelines, AI agents, MCP-based tool orchestration, and agentic workflows.
Design and deploy guardrails, sandboxes, runtime controls, and AI gateway capabilities, including prompt and response inspection, data loss prevention, rate limiting, and circuit breaking.
Build or implement ways to track and govern AI identities, including agents, service principals, workload identities, and delegated access, so CDW always knows what access every AI system has.
Partner with Identity Engineering teams to align AI identities to CDW's enterprise lifecycle, ownership, and least privilege standards.
Partner with Detection Engineering and Security Operations to build detections, telemetry, and response playbooks for prompt injection, jailbreaks, data leakage, model abuse, and tool misuse.
Work with engineering, data science, product, and infrastructure teams to build security in from design through production and publish practical guidance that makes the secure path the easy path.
Present risk assessments, recommendations, and investment priorities to senior leadership and security governance bodies in clear, business-focused terms.
Mentor security engineers and architects to grow AI security expertise across CDW.