Latest in AI
Anthropic Residency 2026: Deadline, Pay and Who Can Apply
The paid research program targets experienced technical leaders, with no prior machine-learning background required.
Anthropic announced open applications for its paid Residency on November 2, inviting established technical leaders into AI safety research. Placements last six weeks to six months, and applications close November 22. Prior machine-learning experience is not required, but this is a full-time research transition rather than an introductory course.
Who the Anthropic Residency is for
The official announcement names quantitative academics, technical founders and staff-level engineers as examples of its intended applicants. A PhD or published papers are not prerequisites.
Residents receive a scoped project, mentorship, access to frontier models and a compute budget. Research areas span safety and alignment, security and frontier red-teaming, and the systems behind model training. The program also includes time and guidance for learning the relevant ML foundations.
This round adapts the earlier Fellows Program for established technical leaders. The distinction is useful: applicants are being asked to bring evidence of judgment and technical execution from another field, rather than arrive with an identical research background.
Deadline, compensation and location
The US role listing gives the application deadline as November 22, 2026 at 11:59 p.m. Pacific Time, with offers expected in January 2027.
Its compensation range is $350,000 to $500,000 a year, prorated for the placement, plus benefits. That is an annualized rate, not a promise that a six-week resident receives the full annual amount.
The role requires full-time commitment. San Francisco is preferred, although remote participation is possible, particularly for shorter placements, from a country where Anthropic can hire. The residency-specific terms describe visa sponsorship for placements lasting at least three months. Candidates should confirm their own location and eligibility with recruiting rather than read this as unrestricted worldwide remote work.
A separate UK listing is linked from the program announcement. Applicants should use the relevant listing rather than assume that US compensation and employment terms apply everywhere.
What an application needs to show
The US process includes an application, recruiter screen, technical assessment, research discussion and references. The form asks for notable work, the applicant's contribution to it, and the research they would most like to pursue.
For an experienced engineer, that suggests a more useful preparation exercise than collecting AI terminology: choose a problem you understand well, explain how you would investigate it, and make your own contribution legible. A good account of a difficult technical decision says more than a broad claim of enthusiasm.
The program's stated aim is to help residents continue in full-time AI research, at Anthropic or elsewhere. It is not a guaranteed permanent job. The interesting signal is that a frontier lab is explicitly treating experience from other technical fields as something research teams can build on.
I'm Paras. I work at the intersection of AI, engineering and distribution.
I help technical founders turn complex work and research into clear stories that developers and founders can understand and share.