Anthropic AI Safety Fellow, UK

Details

Anthropic’s mission is to create reliable, interpretable, and steerable AI systems that are safe and beneficial for users and society. The team includes researchers, engineers, policy experts, and business leaders collaborating to build beneficial AI systems.

Responsibilities

• Participate in the Anthropic Fellows Program, an external collaboration initiative accelerating AI safety research by providing promising talent with research experience.
• Work on empirical projects aligned with AI safety research priorities using external infrastructure like open-source models and public APIs.
• Produce public outputs such as paper submissions with substantial support including mentorship, funding, compute resources, and shared workspace access.
• Collaborate under mentorship from Anthropic researchers and connect with the broader AI safety research community.

Requirements

• Motivated to reduce catastrophic risks from advanced AI systems.
• Excited to transition into full-time empirical AI safety research with potential interest in a full-time role at Anthropic (no guarantee of offers).
• Strong technical background in computer science, mathematics, physics, or related fields.
• Programming skills, particularly in Python and machine learning frameworks.
• Availability to work full-time for at least 2 months, ideally 6 months.
• UK-based or able to work remotely from the UK with a preference for London.
• Eligible to work in the UK; visa sponsorship is not available for Fellows, but F-1 visa holders eligible for OPT/CPT can participate.
• Able to work independently and incorporate feedback.
• Thrive in fast-paced, collaborative environments.

Strong candidates may also have experience in empirical ML research, working with Large Language Models, research areas like interpretability, deep learning frameworks, experiment management, or open-source contributions.

No formal certifications or 100% skill match required.

Interview Process

Includes application submission, a 90-minute Python coding screen, technical interviews without ML focus, final interviews with research discussions, a take-home project, and reference calls.

Benefits

• Weekly stipend of £1,300 for 40 hours per week.
• Access to medical, dental, and vision insurance benefits (varies by country).
• Funding for compute and research expenses.
• Shared workspaces in Berkeley, California and London, UK.
• Employment via a third-party talent partner.

Role-Specific Location Policy

• Remote work is allowed anywhere in the UK with a strong preference for London presence at the shared workspace.
• Exempt from usual in-office time policies.

Additional Information

The program aims to onboard a new cohort starting October 2025, with rollouts continuing later if needed. The fellowship duration can extend up to 6 months based on progress and collaborations.

Anthropic values diversity and inclusion and encourages applications from underrepresented groups in tech.