Sr. Security Engineer, Operations

Overview

We are looking for a senior engineer to solve difficult security problems across Microsoft AI.

Microsoft AI develops and trains first-party models that Microsoft publishes and uses across products including Foundry, Copilot, Microsoft 365, and MDASH. The work spans model development and serving, large-scale compute, shared infrastructure, engineering systems, identity, data, and productivity services - the whole operational system that makes large-scale training, experimentation, and deployment possible. In the last year, MAI evolved from a research lab to a full production model factory. This role is designed for an engineer who has independently built and operated production engineering or security capabilities and is prepared to do so again while the frameworks, operating mechanisms, and secure defaults for frontier AI infrastructure are still being built.

You may come from product or software engineering, security operations, or infrastructure engineering, including detection and response work within those areas. What matters most is builder depth and ability to apply practical risk judgment, not which path you took. You should be able to work through an unfamiliar problem using available evidence, take ownership of a meaningful piece of work, and carry it to a completed, validated result.

This is a broad security-engineering role by design. Initial work may involve building an automated control into our agent framework, improving identity or secrets handling in training systems, defining end-to-end instrumentation and detection for a service, investigating an attack path, hardening a developer workflow, prototyping an AI-assisted security capability, or helping a platform team resolve a systemic risk. You are not expected to begin as an expert in every domain. You will write code, analyze systems and data, build tools and integrations, test controls, and work with engineering teams through implementation and validation. You will use AI routinely for research, coding, analysis, testing, investigation, and automation while remaining responsible for the correctness, security, limitations, and maintainability of the output.

We earn trust with the MAI research community by staying close to the systems we secure, shipping useful improvements rather than only raising findings, making our reasoning and tradeoffs explicit, and telling partners the true state of a risk even when it is inconvenient. The goal is faster, safer models and product delivery, not a new bottleneck.

What You Will Do

Own Security Engineering Problems End to End

• Take complex security problems from discovery and problem framing through design, implementation, deployment, adoption, validation, measurement, and iteration.

• Use architecture analysis, threat modeling, code and configuration review, telemetry, adversarial testing, incident evidence, and data analysis to identify material risks and practical intervention points.