YourNext
Opportunity.org

Astra Fellowship Strategy and Governance Stream: $8,400 Monthly Stipend | AI Safety Research

Category
Fellowships
Posted
September 16, 2026
Region
North America
Country
United States
Sector
Technology & AI
Work mode
On-site
Apply via
bit.ly
Apply nowon bit.ly →

Overview

The Astra Fellowship Strategy and Governance Stream is a fully funded, in-person fellowship for experienced researchers and professionals who want to reduce catastrophic risks from artificial intelligence. Fellows are paired with senior mentors and work full-time on AI safety research from September 2026 to February 2027. More than 80% of the first Astra cohort went on to full-time AI safety roles, and earlier Strategy and Governance fellows have worked on AI strategy research, AI verification discussions with international stakeholders and research on trends in AI progress.

Research areas

Most fellows work with a mentor on a topic tied to that mentor's priorities. Mentor organisations include Forethought, AI Policy Network and Coefficient Giving, along with independent mentors building new organisations. Possible topics include macrostrategy for AI safety, preventing excessive concentration of power, exploring better-future scenarios and implementing AI strategy within governments or leading AI companies. Independent research places are limited and offered for topics considered especially important.

What you get

Fellows receive a stipend of $8,400 per month and a $2,000 travel budget for workshops and conferences. They have weekly meetings with a senior expert and with a research manager, plus career support from a talent mobilisation team and startup incubation help with operations, communications, hiring and fundraising. The main location is the Constellation Institute in Berkeley, California, with a secondary hub for UK-based fellows at the London Initiative for Safe AI. Remote participation may be considered in rare cases, and visa support is available for eligible J-1, F-1/CPT or OPT arrangements.

Who can apply

The stream seeks applicants with deep familiarity with catastrophic AI risks who communicate clearly, reason through complex issues, work independently and develop their own research agendas. Experience in macrostrategy research, grantmaking, cause prioritisation, policy or empirical AI research is preferred, along with understanding of key stakeholders such as frontier AI companies, governments and investors.

How to apply

Applicants submit a resume and at least two references, with writing samples highly recommended. Selection has three stages: initial screening, possible mentor-specific screening with work tests and a referral check, and a final interview. Interested people can register their interest to hear when the next cohort launches. The source gives no application deadline.

Apply nowon bit.ly →

Explore more: opportunities in United States · Technology & AI opportunities