Anthropic Fellows Program — ML Systems & Performance

About Anthropic

Anthropic's mission is to create reliable, interpretable, and steerable AI systems. The organization is committed to building safe and beneficial AI for users and society. Our rapidly growing team comprises researchers, engineers, policy experts, and business leaders collaborating to advance AI safety and performance.

Program Overview

The Anthropic Fellows Program is designed to foster exceptional AI research and engineering talent through direct funding and mentorship. This 4-month, full-time fellowship provides promising technical professionals—regardless of prior experience—the opportunity to work on empirical projects aligned with Anthropic's research priorities.

Fellows leverage external infrastructure including open-source models and public APIs to develop research that produces public outputs, such as peer-reviewed paper submissions. Historical data shows over 80% of fellows from previous cohorts successfully published research outcomes.

The ML Systems & Performance Workstream

This specialized track focuses on advancing machine learning systems performance, optimization, and interpretability. Candidates should have strong technical foundations in machine learning, systems design, and empirical research methodologies.

What You'll Receive

• Compensation: Weekly stipend of $3,850 USD / £2,310 GBP / $4,300 CAD + country-specific benefits
• Research Budget: Approximately $15,000/month in compute funding and research expenses
• Mentorship: Direct guidance from Anthropic's senior researchers
• Workspace: Access to shared workspace in Berkeley, California or London, UK
• Community: Connection to the broader AI safety and security research community
• Duration: 4 months of immersive, full-time research engagement

Interview & Selection Process

The hiring process includes three key stages: initial application review with reference checks, technical assessments and interviews, and a dedicated research discussion. Anthropic encourages applications from all candidates who are genuinely interested in this work, even if you don't meet every qualification. Research shows underrepresented groups in tech often underestimate their candidacy—we urge you to apply if you're passionate about AI safety and performance research.

Key Details

• Next Cohort Start Date: July 20, 2026
• Application Deadline for July Cohort: April 26, 2026
• Rolling Applications: Later cohorts reviewed on a rolling basis; exceptional candidates may start outside standard timelines
• Program Structure: Structured research fellowship with clear deliverables and publication goals

This fellowship is ideal for researchers and engineers passionate about advancing machine learning systems, AI safety, and interpretability. You'll work alongside world-class researchers on problems with significant social and ethical implications.

💰 This position includes a disclosed weekly stipend ($3,850 USD equivalent). Annual equivalent estimate for comparison: approximately $200,000 USD, though this is a time-limited 4-month fellowship with additional compute funding.