Anthropic Fellows Program Is Now Open — Here’s Your Chance to Shape the Future of AI Safety
The Anthropic Fellows Program is now accepting applications for its next cohort, and the window is closing fast, the deadline is April 26, 2026. For researchers, engineers, and emerging technologists with a passion for building safer AI systems, this could be a defining career opportunity.
Anthropic, the public benefit corporation behind the Claude family of AI models, launched the Anthropic Fellows Program as a structured, full-time research initiative designed to grow the next generation of AI safety talent. The program is intentionally built for people with strong potential not just polished CVs. If you have the drive and the technical curiosity, Anthropic wants to hear from you.
The fellowship runs for four months at approximately 40 hours per week, with the next cohort kicking off on July 20, 2026. Fellows receive a weekly stipend of $3,850 (or £2,310 / CAD $4,300), along with up to $15,000 per month in compute and research infrastructure funding. Workspace access is available in Berkeley, California and London, United Kingdom, with remote participation also on the table for eligible candidates in the US, UK, and Canada. Visa sponsorship, however, is not offered.
What makes the Anthropic Fellows Program stand out is its track record, more than 80% of past fellows have produced published research papers, a number that speaks to how seriously the program takes scientific output. Fellows are expected to work on empirical projects using open-source models and public APIs, contributing to real, publishable work in AI alignment and safety.
The program is organized across five workstreams: AI Safety, AI Security, ML Systems and Performance, Reinforcement Learning, and Economics and Societal Impacts. Whether your background is in penetration testing, economics, distributed systems, or interpretability research, there is a track designed to put your skills to work on problems that genuinely matter. Across all tracks, strong Python skills and an ability to thrive in fast-moving research environments are essential.
Anthropic frames AI safety as an empirical science, rigorous, collaborative, and urgent. The Anthropic Fellows Program reflects that philosophy by pulling together researchers across disciplines to tackle scalable, real-world alignment problems. From adversarial robustness and model interpretability to labor market impacts of AI adoption, the scope of work is both broad and consequential.
Applications are reviewed on a rolling basis, so early submission is strongly encouraged. Candidates can apply directly on Anthropic’s official portal. The original listing was also reported by Global South Opportunities, which tracks global research and fellowship openings.
