AI and Society Fellowship – Center for AI Safety

Program overview

The fellowship is a fully-funded, three-month program in which fellows work with significant autonomy — defining and pursuing their own research directions — while drawing on our internal expertise and network of partners for grounding, feedback, and collaboration. By the end of the program, fellows will produce initial results that can be further developed into a publicly shareable format, such as a journal article, blog post, or lecture series. The program will conclude with a final workshop where fellows present their research to fellow cohort members and invited external academics.

The program runs June 1 – August 21, 2026, and will take place in-person at our offices in San Francisco.

The CAIS community

Our researchers have published papers in leading journals including Nature and top AI conferences such as NeurIPS and ICML, while our Executive Director advises on AI safety for leading organizations. Throughout the summer, fellows will have the opportunity to take part in events with researchers from CAIS and external institutions in the Bay Area. The fellowship aims to provide sustained exposure to frontier AI developments, peer communities focused on AI as a central issue, and institutional space to pivot research agendas toward AI.

Key program details

• Dates:
• June 1 – August 21, 2026
• Location:
• San Francisco, California
• Fellowship stipend:
• $25,000
• Covered expenses:
• Travel to and from San Francisco
  Lunch and dinner provided daily at the office
• Events:
• Regular guest speaker events, workshops, and social gatherings with academics and AI researchers throughout the program.

Eligibility

We welcome applications from scholars who meet the following criteria:

• Education and experience:
• One of the following:
  1. Professors (Full, Associate, or Assistant)
  2. Others with a PhD/JD, or ongoing PhD/JD students
  3. Researchers with PhD/JD-equivalent experience.
• Field:
• Economics, international relations, law, or an adjacent discipline such as political philosophy, security studies, etc. If you're unsure whether your background qualifies, email fellowships@safe.ai.
• Research background:
• A demonstrated record of producing high-quality research.

Application process

Applicants will submit a CV, writing sample, brief research proposal, and goals for the fellowship. A select number of applicants will then be invited to interview remotely, and final decisions will be communicated to all applicants by mid-April.