Research Intern - AI Safety, Alignment, and Post-Training

Microsoft•Redmond, WA
•$6,810 - $14,610•Onsite

About The Position

Microsoft’s Research Internships provide a dynamic environment for research careers, with access to research labs led by scientists and engineers working across scientific and technical disciplines. This group studies alignment and safety in frontier AI systems, connecting evaluation, automated red teaming, and post-training. Its work includes developing evaluations to surface risks such as deception, sandbagging, misalignment, and unsafe autonomous behavior; building automated red-teaming systems to stress-test models and agents in realistic environments; and developing post-training methods to improve safety without sacrificing capabilities. Interns work on open research problems in frontier evaluations, red teaming, alignment science, and safety post-training. Over the 12-week internship, they are paired with mentors, collaborate with other interns and researchers, present findings, and contribute to the research community. Interns are expected to be physically located at their manager’s Microsoft worksite for the internship. The position requires a cover letter, relevant work or research samples, and at least two reference letters.

Requirements

  • Currently enrolled in a doctorate degree program in a relevant field, such as Computer Science, Machine Learning, Artificial Intelligence, or a related technical discipline.
  • Must be physically located at the manager’s Microsoft worksite for the duration of the internship.
  • Submit a cover letter, relevant work or research samples, and at least two reference letters.

Nice To Haves

  • Demonstrated research background in machine learning and foundation models.
  • Experience conducting empirical research with large language models.
  • Familiarity with AI alignment, frontier evaluations, red teaming, or post-training.
  • Demonstrated programming, experimentation, and quantitative analysis skills.
  • Research experience studying alignment and safety of frontier AI systems.
  • Experience developing novel evaluations or automated red-teaming methods for advanced models and agents.
  • Experience with post-training methods, including reinforcement learning, supervised fine-tuning, or preference-based methods.
  • Interest in empirically studying emerging behaviors such as deception, sandbagging, reward hacking, and agentic misalignment.

Responsibilities

  • Own a scoped research project in AI alignment, frontier evaluations, red teaming, or safety post-training.
  • Design and run experiments on frontier models and agentic systems.
  • Build evaluations and automated red-teaming approaches to identify emerging safety risks.
  • Develop or evaluate post-training techniques intended to improve model behavior.
  • Collaborate with researchers and engineers to turn experimental findings into scientific insights, publications, and improvements to deployed AI systems.
  • Present findings and collaborate with mentors, other Research Interns, and researchers during the internship.
© 2026 Teal Labs, Inc
Privacy PolicyTerms of Service