Anthropic Fellowship: Eligibility, Funding and How to Apply
By Google News AI ·
Written and maintained by CASRAI Editorial Board
Short answer: the Anthropic Fellows Program is a full-time, four-month, funded research placement run by Anthropic for people who want to work on AI safety questions. As described on the official pages on 2026-11-07, it does not require a PhD, prior machine-learning experience or published papers, but it does require fluent Python, full-time availability and work authorization in the UK, US or Canada.
This guide reconciles what the program page and the job posting each say, because they differ on application details.
Ask about this guide — your first 2 questions are free.
Suggested question: “How do I apply for the Anthropic Fellows Program?”
The program is run by Anthropic. According to the Anthropic Fellows Program page (read 2026-11-07), fellows receive funding for compute and close mentorship from Anthropic researchers, and work for four months on empirical research questions aligned with Anthropic’s research priorities. Mentors pitch project ideas and fellows choose and shape the project they work on.
The page reports outcomes for the program since early 2025: more than 200 fellows have participated, over 65% have published work from their fellowship, and over 35% of fellows from past cohorts subsequently joined Anthropic full-time to work on AI safety. These are Anthropic’s own figures and are not independently audited.
The page lists scalable oversight, adversarial robustness and AI control, model organisms, mechanistic interpretability, AI security, and model welfare.
The program page states: “You don’t need a PhD, prior ML experience, or published papers,” and that Anthropic cares much more about ability to execute on research than credentials. It says past fellows have come from physics, mathematics, computer science and cybersecurity.
Strong candidates are described as having coding skill, motivation to reduce catastrophic AI risk, and the ability to learn quickly and drive projects under uncertainty.
The job posting (read 2026-11-07) adds the hard requirements:
The program page gives compute funding of about 15,000 USD per month and says stipend details are in the job description. The job posting lists a weekly stipend of 3,850 USD, 2,310 GBP or 4,300 CAD depending on country, plus benefits that vary by country, and states that compute and research expenses are funded.
Workspaces named in the posting are London and Berkeley (Toronto is also listed), with remote options in the UK, US and Canada.
The program page directs applicants to the job posting for the application process. The posting describes this sequence: an application and reference check, then technical assessments and interviews, and a research discussion. The posting also says applications go through Constellation rather than the Greenhouse form.
On 2026-11-07 the program page said applications are rolling and to apply by October 18 to be considered for the January cohort, while the job posting said to apply by October 30 for the January cohort. Because cohort deadlines are specific to each intake and the two pages disagreed, treat any date on a third-party site as unconfirmed and check both official pages before planning around a deadline.
No date for any later cohort is stated here.
For other AI-safety and fellowship funding described on CASRAI, see these guides. They are different programs with their own eligibility rules, so compare each on its own official page.
For the wider landscape of award…
Originally published by Google News AI
