AI Security Researcher II

Microsoft•Redmond, WA

1d•Remote

About The Position

Microsoft Security is dedicated to making the world a safer place by addressing critical customer priorities in a landscape of digital threats, regulatory scrutiny, and estate complexity. We aim to revolutionize security and equip every user, customer, and developer with a comprehensive security cloud. Our organization accelerates Microsoft’s mission to secure digital technology platforms, devices, and clouds in diverse customer environments, as well as our own internal systems. We foster a culture of growth mindset, inspire excellence, and encourage our teams and leaders to perform at their best daily, driving innovations that impact billions globally. This role is for an AI Security Researcher II within Microsoft’s AI Red Team research group, focusing on the security of AI agents. The ideal candidate will have research experience in AI security, adversarial machine learning, security for AI, and/or the study of agentic AI systems. You will investigate the failures of autonomous, tool-using, and multi-agent systems and develop methods to improve their security. Your work will involve framing research questions, designing experiments, analyzing risks in agentic systems, and translating findings into practical techniques for Microsoft’s AI products. You will be part of an interdisciplinary team of researchers, red teamers, and engineers dedicated to proactively understanding and mitigating security risks in Microsoft’s major AI systems, including Phi series, Bing Copilot, Security Copilot, GitHub Copilot, Office Copilot, Windows Copilot, and Azure OpenAI. Remote work is an option for this position.

Requirements

Bachelor's Degree in Statistics, Econometrics, Computer Science, Electrical or Computer Engineering, or related field AND 2+ years related experience (e.g., statistics, predictive analytics, research) OR Master's Degree in Statistics, Econometrics, Computer Science, Electrical or Computer Engineering, or related field AND 1+ year(s) related experience (e.g., statistics, predictive analytics, research) OR Doctorate in Statistics, Econometrics, Computer Science, Electrical or Computer Engineering, or related field OR equivalent experience.
Ability to meet Microsoft, customer and/or government security screening requirements are required for this role.
Microsoft Cloud Background Check: This position will be required to pass the Microsoft background and Microsoft Cloud background check upon hire/transfer and every two years thereafter.

Nice To Haves

Demonstrated research output in AI security, adversarial machine learning, agentic systems, LLM evaluation, or related areas, such as publications, preprints, technical reports, talks, open-source tools, or research artifacts
Experience designing experiments, evaluations, benchmarks, or methodologies to study security-relevant behavior in AI systems, especially autonomous, tool-using, or multi-agent systems
Ability to work collaboratively in an interdisciplinary team environment
Ability to connect research findings to practical security guidance, mitigations, or evaluation methods for real-world AI products

Responsibilities

Formulate and investigate research questions at the intersection of cybersecurity and agentic AI behavior, including security risks, failure modes, and adversarial behavior in agentic AI systems.
Design rigorous experiments, evaluations, and benchmarks for autonomous, tool-using, and multi-agent systems.
Analyze results and develop research insights, methodologies, and mitigations that advance the security of AI systems.
Collaborate with researchers, red teamers, and product teams to translate AI security research into practical impact for Microsoft’s AI products.