Anthropic Residency opens applications, aimed at senior technical leaders

Anthropic has opened applications for the Anthropic Residency, a paid research program lasting 6 weeks to 6 months, aimed at people who have already made significant contributions in technical fields and wish to transition to full-time AI safety research; this round of applications closes on November 22, 2026. The program does not require a machine learning background, a PhD, or published papers, and is intended for professors in quantitative fields, technical founders, and staff-level engineers. Residents will conduct empirical research on questions posed by Anthropic researchers, receive regular one-on-one mentoring from an assigned mentor, and have access to frontier models and a dedicated compute budget, with compensation prorated from an annual salary of 350,000 to 500,000 USD. The preferred location is the San Francisco office, and short-term residencies may be remote. Areas of focus include scalable oversight, adversarial robustness and AI control, misaligned model organisms, model internals and interpretability, guardrail research, AI welfare, as well as frontier red teaming and ML systems and reinforcement learning.