OverviewSecurity represents one of the most critical priorities for our customers amid rapidly evolving digital threats, regulatory scrutiny, and estate complexity. The Cloud & AI organization accelerates Microsoft's mission to secure digital technology platforms, devices, clouds, and AI systems across customers' heterogeneous environments and Microsoft's own internal estate. Our culture centers on growth mindset, inspiring excellence, and helping teams and leaders bring their best every day.
Microsoft Red Team (MRT) emulates real-world advanced persistent threats against Microsoft and an ecosystem of external customers. AI is now central to that mission on both sides of the fight: what red teamers can do today, and what sophisticated threat actors will be able to do tomorrow. Within MRT, Strategic AI advances offensive AI security research and translates emerging risks into practical protection for Microsoft.
As a Member of Technical Staff (Senior AI Security Researcher), you will post-train open-weight language models for cybersecurity tasks, including using recursive self-improvement. You will develop and curate training datasets, define training objectives, implement post-training methods, and run reproducible experiments. You will evaluate how training choices affect task performance, generalization, and security, and deliver trained model artifacts and reusable workflows for red-team partners.
Microsoft's mission is to empower every person and every organization on the planet to achieve more. As employees we come together with a growth mindset, innovate to empower others, and collaborate to realize our shared goals. Each day we build on our values of respect, integrity, and accountability to create a culture of inclusion where everyone can thrive at work and beyond.
Responsibilities- Own hands-on post-training research projects for open-weight language models used for cybersecurity tasks.
- Develop and curate training data, objectives, and experimental controls; assess supervised, preference-based, and reinforcement-learning methods when relevant.
- Implement reproducible training pipelines and track datasets, configurations, checkpoints, and evaluation results.
- Develop and evaluate recursive self-improvement methods for training
- Deliver trained model artifacts and reusable workflows that red-team partners can evaluate and apply.
Qualifications
Required Qualifications:
- Bachelor's Degree in Statistics, Econometrics, Computer Science, Electrical or Computer Engineering, or related field AND 4+ 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 3+ years related experience (e.g., statistics, predictive analytics, research)
- OR Doctorate 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 equivalent experience.
-
Other Requirements:
- Ability to meet Microsoft, customer and/or government security screening requirements are required for this role. These requirements include, but are not limited to the following specialized security screenings:
- Microsoft Cloud Background Check: This position will be required to pass the Microsoft Cloud Background Check upon hire/transfer and every two years thereafter.
Preferred Qualifications:
- Prior experience training open-weight language models with distributed or multi-GPU infrastructure.
- Prior experience with preference optimization or reinforcement-learning-based post-training.
- Prior experience evaluating language models in adversarial or security-focused settings.
- Peer-reviewed publications or released research software in language-model training, post-training, or AI security.
- 2+ years of experience with Python and machine learning frameworks.
- Experience fine-tuning or post-training language models.
- Prior experience post-training language models for cybersecurity tasks or evaluating language models in adversarial or security-focused settings.
- Prior experience developing or applying recursive self-improvement methods for model training, or peer-reviewed publications or released research software in language-model training, post-training, or AI security.
Applied Sciences IC4 - The typical base pay range for this role across the U.S. is USD $119,800 - $234,700 per year. There is a different range applicable to specific work locations, within the San Francisco Bay area and New York City metropolitan area, and the base pay range for this role in those locations is USD $160,200 - $261,000 per year.
Certain roles may be eligible for benefits and other compensation. Find additional benefits and pay information here:
https://careers.microsoft.com/us/en/us-corporate-pay
This position will be open for a minimum of 5 days, with applications accepted on an ongoing basis until the position is filled.
Microsoft is an equal opportunity employer. All qualified applicants will receive consideration for employment without regard to age, ancestry, citizenship, color, family or medical care leave, gender identity or expression, genetic information, immigration status, marital status, medical condition, national origin, physical or mental disability, political affiliation, protected veteran or military status, race, ethnicity, religion, sex (including pregnancy), sexual orientation, or any other characteristic protected by applicable local laws, regulations and ordinances. If you need assistance with religious accommodations and/or a reasonable accommodation due to a disability during the application process, read more about requesting accommodations.
— Sarah J.
·
Marketing Manager
★★★★★
Verified
