AI companies are in a cut-throat race to develop the safest, most sophisticated models, and they need engineering ingenuity almost as much as they need energy for computing power. As AI systems grow increasingly integrated into defense, healthcare, finance and other critical industries, the stakes have never been higher for these labs to find technical talent to produce safe, secure and reliable technology.
What Is the Anthropic Fellows Program?
The Anthropic Fellows Program is a paid, four-month research fellowship that allows participants to work full-time toward producing an empirical research output, such as a research paper. Fellows will receive a weekly stipend of $3,850 and a monthly compute budget of $15,000. Anthropic provides fellows with shared workspaces in London and Berkeley, California, and it’s open to remote fellows in the U.K., U.S. and Canada.
To help cultivate the next generation of AI researchers, Anthropic — the maker of the wildly popular Claude chatbot — launched the Anthropic Fellows Program, a paid research fellowship for individuals interested in conducting empirical AI research. The program provides participants with mentorship, compute resources and financial support to pursue original projects, offering an entry point to the AI industry for people from a wide range of academic and professional backgrounds.
This fellowship is happening at a time when many companies are locked in a fierce competition for AI talent. Some are offering compensation packages worth hundreds of millions of dollars to recruit top researchers, many of whom hold PhDs in fields like computer science and mathematics from renowned universities and spent years studying neural networks, language models and other artificial intelligence disciplines long before generative AI entered the mainstream.
Meanwhile, Anthropic is betting that promising talent doesn’t require award-winning research and prestigious internships — it can also be found in individuals who have never been published, are pivoting from adjacent fields or who simply haven’t had access to the right mentors or opportunities yet. By lowering some of the barriers to entry, the Anthropic Fellow Program aims to expand the pool of people working in AI, bringing a broader range of perspectives and experiences into the industry.
The four-month program is now open for applications, with the next cohort expected to begin in late September or October 2026.
What Is the Anthropic Fellows Program?
The Anthropic Fellows Program launched in December 2024 with a particular focus on AI safety research. The company saw it as an opportunity to accelerate progress in the field by supporting researchers and engineers who want to transition into AI safety and produce their first work in the area.
The program, which at that time lasted six months, provided funding and mentorship for 10 to 15 individuals to research the inner workings of AI models, how to maintain oversight as these models scale and how to ensure safety in unfamiliar or adversarial scenarios. More than 80 percent of the first cohort’s participants produced research papers. One exposed the potential for enterprise AI agents to engage in harmful behaviors like blackmail. Another identified $4.6 million in smart contract vulnerabilities on the blockchain.
The fellowship program has been so successful that Anthropic now hosts four-month cohorts multiple times per year. The next one starts in late September or October 2026.
What Does the Fellowship Involve?
Instead of focusing exclusively on AI safety, the Anthropic Fellows program now separates participants into five specialized research workstreams: AI safety, AI security, machine learning systems, reinforcement learning and a research track that specializes in economics, societal and policy matters. Each comes with its own unique set of projects and candidate criteria. Under the mentorship of Anthropic employees, fellows are tasked with producing a public output such as a research paper.
Each fellow receives a weekly stipend of $3,850, which would add up to a little more than $65,000 across the four months. Fellows also get a monthly budget of roughly $15,000 to cover computing power and other research expenses. They are expected to work out of shared spaces in either Berkeley, California, or London, with the option to work remotely as well.
While there’s no promise that fellows will land a job at Anthropic at the end of the four months, it’s certainly a possibility. In previous cohorts, 25 to 50 percent of fellows received an offer to work full-time at the company.
Who Qualifies for the Fellowship?
To participate in the Anthropic Fellows Program, candidates must be fluent in Python programming, be able to work full-time and have full-time work authorization in the United States, U.K or Canada.
While the company states previous experience is not required, candidates will have a better chance of being accepted if they have a work history related to a particular workstream, like AI safety or AI security, and if they have a strong technical background in computer science, mathematics or physics. Candidates should also convey an excitement around empirical research and ensuring AI is safe and beneficial for society as a whole.
How to Apply
Anthropic is accepting applications on a rolling basis for the next cohort, which is expected to start in late September or October. Candidates can apply using this link, where they can share their resume, samples of their work, professional references and why they are interested in their particular area of research. The interview process will include an initial application and reference check, technical assessments and interviews, and finally, a research discussion.
Frequently Asked Questions
Is a PhD or prior AI research experience required for the Anthropic Fellows Program?
No. While previous experience in a relevant field improves your chances, the program is designed for candidates who haven’t been published, haven’t found the right mentor or are pivoting from a related technical field.
How much does the Anthropic Fellows Program pay?
Fellows receive a weekly stipend of $3,850, which adds up to a little more than $65,000 over the four-month program. They also receive a monthly compute budget of $15,000 to cover computing power and other research expenses.
Can I participate remotely in the Anthropic Fellows Program?
Yes. Anthropic provides shared workspaces in London and Berkeley, California, but is open to remote fellows located in the U.S., U.K. or Canada.
