Fellowship · PRISM Research

PRISM AI Safety Research Fellowship

16-week remote AI safety research fellowship producing peer-reviewed work in interpretability, governance, evaluation and alignment

closed for 2026 / next cycle yes remote

Program overview

Good fit for transitioning into publishable AI safety research; eval/alignment themes match resume, but current cycle passed

Notes

Exact 2026 dates found from PRISM organizer posts/listings; official site is mostly JS and only confirms 16-week fellowship

View official program ↗