Job Opportunities API

THE POSITION Our roster has an opening with your name on it As a Staff AI Security Engineer, you'll drive how FanDuel builds and uses AI securely, from the engineers shipping production agents and models to the business use cases they power. You'll turn security into the path of least resistance and catch risk before it reaches production, across FanDuel's AI gateway, agent orchestration layer, and the growing MCP tool ecosystem.   This role sits within Security and works in close, day-to-day partnership with the teams building and operating FanDuel's AI platform. You'll help shape platform security standards, red team the agents and reference workflows running on it, and make sure security controls are built into the platform itself as it matures, rather than left to individual application teams.   Success in this role means AI/LLM and agentic security is treated as a property of the platform, enforced at the gateway and orchestration layer, not something each team reinvents.   In addition to the specific responsibilities outlined above, employees may be required to perform other such duties as assigned by the Company. This ensures operational flexibility and allows the Company to meet evolving business needs. THE GAME PLAN Define the AI security roadmap:  Own the multi-year AI and agentic security roadmap for FanDuel—the vision, strategy, and standards—developed jointly with the teams building the AI platform, so platform and security standards are one bar, not two.   Secure the inference and agent stack:  Drive AI/LLM and agentic security architecture across the platform as it's built out: the AI gateway and MCP tool registry, the agent orchestration layer, the skills and capability catalogue, and the evals and observability pipeline. Define controls, guardrails, and monitoring for prompt injection, jailbreaking, insecure output handling, training and retrieval data poisoning, model and data supply-chain risk, and excessive agent agency.   Own the MCP and tool-use attack surface:  Define security requirements for MCP server registration, tool authentication and authorization, and agent-to-tool trust boundaries—from tool poisoning and confused-deputy patterns to unauthenticated or over-privileged tool access.   Stand up AI red-teaming:  Build and run AI red-teaming and adversarial testing so failure modes are found before they reach production, and partner with the platform teams to feed those results into the platform's review process.   Own the AI security governance model:  defining assessment frameworks using OWASP LLM Top 10, MITRE ATLAS, and NIST AI RMF to give engineering teams a clear path to shipping AI features safely.   Automate and scale:  Build automation and tooling—automated red-team and adversarial test suites, policy-as-code for the gateway—that turns AI security into a scalable, repeatable, secure-by-default system rather than manual review.   Partner and mentor:  Partner across Security, the AI platform teams, and the business to shape direction and trade-offs. Mentor engineers on both sides and help grow AI security capability across the organization.   THE STATS What we're looking for in our next teammate Deep, hands-on security engineering experience embedded in the software development lifecycle, from design and code review through CI/CD, deployment, and production.   Hands-on AI/LLM and agentic security experience: you understand how these systems fail in practice (prompt injection, jailbreaking, insecure output handling, data and model poisoning, excessive agency, tool-use exploits) and have strong, defensible opinions on how to secure them.   Working knowled