Program Manager, Alignment

San Francisco, United StatesHybrid

Responsibilities

- Drive practical execution across research, engineering, policy, product, legal, communications, recruiting, and external collaborators as needed.
- Create and maintain lightweight systems that help alignment work scale, such as project trackers, decision memos, operating processes, review loops, and recurring coordination forums.

ABOUT THE TEAM OpenAI’s Alignment team works to ensure that increasingly capable AI systems reliably follow human intent, avoid catastrophic behavior, and remain controllable, auditable, and safe as capabilities scale.
- Map the alignment and AI safety community: understand who is working on what, what people care about, where OpenAI can help, and which collaborations are worth pursuing.
experience in research program management, technical program management, operations, chief-of-staff work, partnerships, recruiting, or related roles in fast-moving environments.
- Have familiarity with the AI safety or alignment ecosystem, or can get up to speed quickly enough to know who to talk to and what the community cares about.
Experience working directly with AI safety, machine learning, or research teams. -
Experience building a new program, process, community effort, or function from scratch. -
Experience with fellowships, research communities, external collaborations, technical governance, evaluations, monitoring, or safety review processes.
About OpenAI OpenAI is an AI research and deployment company dedicated to ensuring that general-purpose artificial intelligence

The team is fast-moving: priorities change as evidence, risks, and organizational needs change.
We need people who can help execute important alignment and safety priorities that don’t otherwise have an owner.
ABOUT THE ROLE We are looking for a Research Program Manager for the Alignment team.
You should expect to move between big, ambiguous questions and very concrete execution: writing a crisp project plan, tracking owners, chasing a decision, running a process, organizing a meeting, helping unblock a researcher, or picking up the operational work nobody else has time to do.
OpenAI is a continuously evolving place, and this role will evolve with it.
Some projects will be highly visible and strategic; others will be small, practical, or unglamorous.
The common thread is doing whatever needs to be done in service of alignment/safety at OpenAI.
If you want a role where no work is below you but where scope can expand quickly, and where good judgment and follow-through matter as much as title or function, this may be a strong fit.
We use a hybrid work model of 3 days in the office per week and offer relocation assistance to new employees. IN THIS ROLE,
YOU WILL: - Own 2-4 ambiguous alignment special projects at a time, from problem definition through execution and follow-through.