Anthropic Fellows Program, AI Safety

at Anthropic

📍 Canada
📍 United States
📍 London, United Kingdom
📍 Berkeley, United States
📍 San Francisco, United States

USD 200,200 per year

MIDDLE

✅ Remote ✅ On-site

Tech Stack
Tag name is followed by "@" symbol and proficiency level value. About proficiency levels:

1-2 — basic awareness. Minimal hands-on experience, and a rudimentary understanding of the technology's purpose;

3-6 — daily use. Comfortable and regular usage, capable of handling common tasks and challenges related to the technology;

7-9 — you are an expert, you can teach others, you know all the pitfalls and tricks;

10 — exceptional knowledge, comprehensive understanding, and adeptness in all aspects of the technology, including advanced problem-solving. Think twice before claiming or demanding such level.

AI @ 3 API LLM Machine Learning @ 3 Mathematics @ 6 Python @ 5 Security

Details

Anthropic’s Fellows Program supports promising technical talent in AI research and engineering, regardless of previous experience. Fellows use external infrastructure, including open-source models and public APIs, to work on empirical projects aligned with Anthropic’s research priorities, with the goal of producing a public output such as a paper submission.

Responsibilities

Conduct full-time empirical AI safety research for four months.
Develop and implement research ideas quickly.
Work with Anthropic mentors through a project selection and mentor-matching process.
Produce a public research output, such as a paper submission.
Participate in technical assessments, interviews, research discussions, and the broader AI safety and security research community.
Potentially work on scalable oversight, adversarial robustness and AI control, model organisms, model internals and mechanistic interpretability, or AI welfare.

Requirements

Fluency in Python programming.
Availability to work full-time for the duration of the Fellows Program.
A strong technical background in computer science, mathematics, or physics.
Ability to work in a fast-paced, collaborative environment and communicate clearly.
Motivation to ensure that AI is safe and beneficial for society.
Candidates may also have experience with empirical machine learning research, large language models, relevant AI safety research areas, cybersecurity, economics, social sciences, or open-source contributions.
Must have full-time work authorization in the United States, United Kingdom, or Canada and be located in that country during the program.

Compensation

Weekly stipend of 3,850 USD, 2,310 GBP, or 4,300 CAD, plus benefits that vary by country.
The expected base stipend is 3,850 USD per week, equivalent to 200,200 USD annually based on 52 weeks.
Approximately 15,000 USD per month in compute and other research expense funding.
Four-month program with possible extension.

Work Arrangement

Shared workspaces are available in Berkeley, California, and London, United Kingdom.
Remote participation is available for fellows located in the United Kingdom, United States, or Canada.
Fellows may be asked about their availability to work from Berkeley or London on a full- or part-time basis.

Visa and Program Information

Anthropic is not currently able to sponsor visas for Fellows. Participants must already have or independently obtain full-time work authorization in the United Kingdom, United States, or Canada.
Anthropic does not guarantee full-time offers after the program.
Applications and interviews are managed by Constellation, Anthropic’s recruiting partner.