Microsoft
AI Security Engineer
$102K–$261K
United States · On-site · Full-time · Posted Oct 8, 2026
- Experience
- Minimum: 1+ year with a master's degree, or 2+ years with a bachelor's degree, in the specified experience areas; or equivalent experience. Preferred alternatives: doctorate; master's plus 3+ years; or bachelor's plus 5+ years; or equivalent experience.
- Education
- Master's degree in Statistics, Mathematics, Computer Science, Computer Security, or related field; or bachelor's degree in those fields or related field; or equivalent experience.
About the job
Implements and runs controlled evaluations of frontier AI models and agentic systems in cyber environments to assess meaningful offensive cybersecurity capabilities and communicate findings.
What you'll do
- Build, maintain, and operate cyber-capability evaluations, tasks, cyber ranges, vulnerable applications, and evaluation harnesses.
- Run controlled model experiments, define success criteria and baselines, collect telemetry, and analyze results and evaluation artifacts.
- Apply sandboxing and containment controls, and communicate evaluation findings through reports, documentation, and briefings.
What they ask for
- Master's degree in a listed or related field and 1+ year of relevant experience; or bachelor's degree in a listed or related field and 2+ years of relevant experience; or equivalent experience.
- Meet Microsoft, customer and/or government security screening requirements, including the Microsoft Cloud Background Check.
- Provide proof of citizenship, U.S. permanent residency, or other protected status for export-control eligibility assessment; the posting also specifies citizenship verification for this position.
Nice to have
- Programming ability, particularly in Python, and experience building automation, scripts, or security tooling.
- Experience with AI or agentic model evaluations, offensive security, controlled experiments, or related security labs and cyber ranges.
Pay
USD $102,100 - $202,200 per year (base pay); USD $133,800 - $219,200 per year (base pay); USD $119,800 - $234,700 per year (base pay); USD $160,200 - $261,000 per year (base pay)
Tools
Python, PyRIT
AI focus
Evaluate frontier AI models and agentic systems for offensive cyber capabilities, analyze model behavior and evaluation artifacts, and improve testing, scoring, telemetry, and containment.
Model-written summary · the original posting is authoritative