Anthropic Fellows Program — AI Safety

Details

Anthropic’s mission is to create reliable, interpretable, and steerable AI systems. The Anthropic Fellows Program is designed to foster AI research and engineering talent by providing funding and mentorship to promising technical candidates to work on empirical AI safety projects with the goal of producing a public output (e.g., a paper submission). The next cohort starts July 20, 2026; apply by April 26, 2026 for that cohort. Applications are reviewed on a rolling basis.

What to expect

• 4 months of full-time research (with possible extension)
• Direct mentorship from Anthropic researchers
• Access to a shared workspace (Berkeley, California or London, UK)
• Connection to the broader AI safety and security research community
• Weekly stipend of 3,850 USD / 2,310 GBP / 4,300 CAD + benefits (vary by country)
• Funding for compute (~$15k/month) and other research expenses

Interview process

The process includes an initial application & reference check, technical assessments & interviews, and a research discussion. The program encourages applicants who may not meet every listed qualification to apply.

Fellows workstreams

Available workstreams include (examples):

1. AI Safety Fellows
2. AI Security Fellows
3. ML Systems & Performance Fellows
4. Reinforcement Learning Fellows
5. Economics & Societal Impacts Fellows

Fellows will primarily use external infrastructure (open-source models, public APIs) to work on empirical projects aligned with Anthropic research priorities. Projects aim to produce public outputs; in earlier cohorts over 80% of fellows produced papers.

AI Safety specifics (mentors & research areas)

Potential mentors and example research areas include:

• Mentors (sample): Sam Bowman, Sara Price, Alex Tamkin, Nina Panickssery, Trenton Bricken, Logan Graham, Jascha Sohl-Dickstein, Joe Benton, Collin Burns, Fabien Roger, Samuel Marks, Kyle Fish, Ethan Perez
• Research areas: Scalable Oversight, Adversarial Robustness and AI Control, Model Organisms, Model Internals / Mechanistic Interpretability, AI Welfare
• Links to past projects and representative posts are provided in the posting for examples of prior Fellows work and research directions.

Requirements

• Must be available to work full-time on the Fellows program (expected 40 hours/week)
• Fluent in Python programming
• Strong technical background in computer science, mathematics, or physics (or similarly relevant experience)
• Work authorization in the US, UK, or Canada and be located in that country during the program
• Experience with empirical ML research projects and/or large language models is strongly relevant
• Track record of open-source contributions is desirable

Logistics

• Workspace locations: shared workspaces in London and Berkeley; Anthropic is also open to remote fellows in the UK, US, or Canada. Fellows will be asked about availability to work from Berkeley or London (full- or part-time) during the program.
• Visa sponsorship: Anthropic is not currently able to sponsor visas for fellows; participants must have or independently obtain full-time work authorization in the UK, the US, or Canada.
• Program duration: 4 months, full-time. Applicants who cannot commit to full duration should still apply and note constraints.

Compensation

• Expected base stipend: 3,850 USD / 2,310 GBP / 4,300 CAD per week (40 hours/week for 4 months)
• Funding for compute (~$15,000 per month) and other research expenses

How to apply

To apply, complete the Constellation application form (application links provided in the posting). Constellation is Anthropic’s recruiting partner for this program and manages applications and interviews.

Additional notes

• The program does not guarantee full-time offers, though strong performance may lead to full-time opportunities. In past cohorts, 25–50% of fellows received full-time offers.
• Applicants should be cautious of scams; Anthropic recruiters contact from @anthropic.com addresses and will not ask for money or banking information before the first day.