A funded AI safety research fellowship run by Anthropic, designed to accelerate AI safety research and develop research talent. Fellows work on safety-relevant research problems with mentorship from Anthropic researchers; cohorts and application windows are announced on the Anthropic Alignment Science blog.