Anthropic Fellows Program

Anthropic Fellows Program

Offers additional fellowship opportunities for AI researchers through collaborative cohorts.

The Anthropic Fellows Program — July 2026 Cohort is the second of two active four-month research fellowships offered by Anthropic in 2026, running on the same structure as the May cohort. Fellows receive a weekly stipend of $3,850 USD (2,310 GBP / 4,300 CAD for UK- and Canada-based participants) and approximately $15,000 per month in cloud compute funding. The program places fellows in close mentorship relationships with Anthropic scientists, and all research outputs are published publicly.

Eligibility mirrors the May cohort: the fellowship is for individual researchers affiliated with universities or research organizations; no PhD, prior machine-learning publications, or established ML track record is required. Anthropic's stated selection bar is strong Python coding ability, the capacity to turn ambiguous problems into concrete research progress, and genuine motivation to address catastrophic AI risk. Historical outcomes from earlier cohorts show more than 80 percent of fellows produced papers and over 40 percent joined Anthropic full-time after completion.

Funded research areas span scalable oversight, adversarial robustness and AI control, model organisms of misalignment, mechanistic interpretability, AI security, and model welfare. The July 2026 cohort shares its application pipeline with the May cohort — applicants submit through Anthropic's Greenhouse board and proceed through reference checks, technical assessments, and a research discussion. Researchers who missed the May cohort deadline should treat the July intake as a second discrete window with its own rolling-review cycle.