Senior AI Model Security Researcher

Microsoft

Confirmed live today High trust

Hybrid

Quick summary

Work type: Hybrid
Location: —
Salary: $119,800–$234,700 / yr
Posted: 1 day ago
Freshness: Confirmed live today
Closes: May 3, 2027

Market check

Salary context

Competitive pay

How this pay compares to similar roles

Similar $191k

This role $177k

$106k $248k

below market most similar roles pay here above market

This role pays less than 56% of similar roles. Most pay $159,000–$223,750 — the blue band above. At the midpoint, this role pays about $177k versus about $191k for comparable roles.

Based on 240 similar postings.

Employer

About Microsoft

Microsoft Corporation is a global technology leader producing software, hardware, and cloud services including Windows, Office 365, Azure cloud platform, Xbox gaming, and Surface devices. Industry: Software & Cloud Computing

Microsoft currently has 676 open roles on FindRole.

Listed pay typically runs $119,800–$234,700 across 630 roles with salary data.

Most-posted roles

View all roles at Microsoft

At a glance

TL;DR · Senior AI Model Security Researcher

Role Posting Log in to save

The Senior AI Model Security Researcher joins a new red team focused on adversarial machine learning, model integrity, offensive security, automation, and privacy. This role involves conducting hands-on red teaming of advanced AI models, training systems, and evaluation environments. You will design white-box attacks to expose exploitable behaviors such as jailbreaks, reward hacking, data poisoning, and privacy leakage. Key responsibilities include building high-fidelity experiments, automated attack suites, and rigorous threat models to produce reproducible evidence. You will partner with scientists and engineers to convert findings into regression evaluations, adversarial training, and stronger technical controls. Required skills include expertise in adversarial machine learning, offensive security, and research prototyping. The work addresses critical security problems involving large language models, post-training systems, reinforcement learning, and model supply-chain security.

Skills

Adversarial Machine Learning AI Security Offensive Security Large Language Models reinforcement-learning Red Teaming Data Poisoning Privacy Attacks Model Evaluation Cyber Exploitation Identity Security Sandboxing Data Pipelines

What you'll do

Design and execute white-box attacks against advanced AI models, data pipelines, and evaluation environments.
Investigate adversarial behaviors including jailbreaks, reward hacking, data poisoning, and privacy leakage.
Build high-fidelity experiments and automated attack suites to reproduce realistic threats continuously.
Develop rigorous threat models and attack hypotheses to assess severity and governance impact.
Partner with engineering teams to identify root causes and drive findings to verified remediation.
Convert confirmed weaknesses into regression evaluations, adversarial training, and stronger technical controls.
Communicate findings and residual risk clearly to technical teams, leadership, and independent reviewers.

What we're looking for

Doctorate in a relevant field OR Master's Degree with 3+ years of research experience OR Bachelor's Degree with 4+ years of research experience.
Equivalent experience to the listed educational and research requirements.
Ability to pass the Microsoft Cloud background check.
Hands-on expertise in adversarial machine learning, AI security, model evaluation, privacy attacks, or offensive security (preferred).
Experience designing experiments or security investigations that produce reproducible technical evidence and actionable remediation (preferred).
Programming and research-prototyping ability to build attack tooling, evaluation harnesses, and automated tests (preferred).
Experience testing large language models, post-training systems, reinforcement-learning systems, agentic systems, or AI-enabled security products (preferred).
Experience with cyber exploitation, cloud or identity security, sandboxing, secure evaluation environments, or offensive-security operations (preferred).

Save