Paid Anthropic Residency

Anthropic, the AI safety company behind Claude, is opening a round of its Anthropic Residency, a paid, fixed-term research programme for people who have already led ambitious technical work and want to move into full-time AI safety research. Residents spend six weeks, three months or six months on a research problem set by Anthropic researchers, with a dedicated mentor, frontier models and their own compute budget. It is built for professors in quantitative fields, technical founders and staff-level engineers, and no machine learning background is required.

What you get:

• Pay of $350,000 to $500,000 a year in the US, or £265,000 to £375,000 a year in the UK, pro-rated for the length of your placement, plus benefits
• One-to-one mentorship from an Anthropic researcher, from a list that includes the leads of the Alignment Science, Scalable Oversight, Alignment Stress-Testing and Model Psychology teams
• Access to frontier models and a dedicated compute budget for your project
• Self-study time, talks and guidance to get up to speed on modern machine learning and AI safety research
• Visa sponsorship for placements of three months or longer

What you'll do:

• Work full-time on an empirical research problem matched to your interests and background, in areas such as scalable oversight, AI control, model organisms of misalignment, interpretability, safeguards, AI welfare, frontier red-teaming or ML systems and reinforcement learning
• Work from the San Francisco office, which Anthropic prefers, or remotely, which is possible especially for shorter placements
• Aim to finish ready to keep working on AI research full-time, at Anthropic or elsewhere

Key dates:

• Applications close November 22, 2026, 11:59 PM Pacific
• Offers expected in January 2027
• Start dates are flexible

More information:

• Apply through the US posting (San Francisco or New York) or the UK posting (remote UK, with the option of the London office), both linked from the programme page
• The interview process includes an application, a recruiter screen, a technical assessment, a research discussion and a reference check
• Applicants may also be considered for the Anthropic Fellows Program

What you get:

• Pay of $350,000 to $500,000 a year (US) or £265,000 to £375,000 a year (UK), pro-rated for the placement, plus benefits
• One-to-one mentorship from an Anthropic researcher
• Frontier model access and a dedicated compute budget
• Self-study time, talks and guidance on modern ML and AI safety research
• Visa sponsorship for placements of three months or longer

What you'll do:

• Work full-time for six weeks, three months or six months on an empirical AI safety or security research problem matched to your background
• Work from San Francisco (preferred) or remotely, especially for shorter placements
• Prepare to continue in AI research full-time, at Anthropic or elsewhere

Key dates:

• Applications close November 22, 2026, 11:59 PM Pacific
• Offers expected January 2027
• Flexible start dates; placements of six weeks, three months or six months

More information:

• Anthropic posts separate US and UK listings on its own jobs board; apply to the one for where you would work
• Applicants may also be considered for the Anthropic Fellows Program

• Have led ambitious work in a technical field, for example as a professor in a quantitative discipline, a technical founder or a staff-level engineer
• Motivated by making sure AI is safe and beneficial for society as a whole
• Excited to transition into empirical AI research full-time
• Able to implement ideas quickly and communicate clearly
• No PhD, prior machine learning experience or published papers required
• Full-time commitment for the length of the placement
• Remote participants must be based in a country where Anthropic can hire
• Minimum education listed on the postings: a bachelor's degree or an equivalent combination of education, training or experience