Anthropic Fellows Program 2026: Shaping the Future of AI Safety

The field of artificial intelligence is growing quickly, creating new chances for people who work in research, engineering, and policy. As more companies and governments use advanced AI, there’s a bigger need for experts who can build AI systems that are safe, easy to understand, and helpful. The Anthropic Fellows Program 2026 is a four-month, fully funded research fellowship designed to help the next generation of AI talent. This program focuses on AI safety, security, learning, governance, and advanced machine learning. It offers mentorship, money, and resources to help fellows work on important AI safety issues with leading researchers. Applications are accepted continuously for groups starting in late September 2026 and later.

About the Anthropic Fellows Program

The Anthropic Fellows Program is a full-time research initiative. It lets participants work on AI research projects that match Anthropic’s main research goals. Fellows get guidance from experienced researchers and engineers. They also help with projects that aim to make AI systems safer and more beneficial. This program is for people who have strong technical or analytical skills but want structured guidance, research experience, and professional exposure in the fast-growing AI field. Many fellows from past programs have created public research, like papers, technical reports, and open-source contributions.

Fellowship Workstreams

Fellows can join different workstreams based on their background and interests. Each workstream tackles important AI problems and helps improve the understanding of safe AI use worldwide.

Available Fellowship Workstreams

1. AI Safety Fellows: This track focuses on reducing risks from advanced AI systems.
2. AI Security Fellows: This track looks into AI-related cyber risks and how to analyze vulnerabilities.
3. Machine Learning Systems and Performance Fellows: This group works on large-scale AI infrastructure and making systems run better.
4. Reinforcement Learning Fellows: Participants here focus on reinforcement learning systems and how AI learns through trial and error.
5. Economics and Policy Fellows: This track studies the effects of advanced AI on society and the economy.

Key Research Areas

Fellows might work on topics such as:

• Ways to oversee advanced AI systems as they grow.
• Methods for aligning AI with human goals and controlling its behavior.
• Making AI systems resistant to attacks and improving AI security.
• Understanding how AI models work internally.
• Ensuring AI benefits humans and considering AI ethics.
• Developing reinforcement learning systems.
• Looking at risks from AI in cybersecurity.
• Creating rules and policies for AI.
• Studying the economic effects of AI.
• Testing and evaluating large language models.
• Finding and fixing AI weaknesses through testing.
• Building and managing large AI systems.
• Training and improving AI models.

Participants will work with mentors who are leaders in AI safety.

Fellowship Benefits

The Anthropic Fellows Program offers significant support, including financial, technical, and professional development.

Financial Support

Selected fellows receive weekly payments based on their location:

• 3,850 USD per week
• 2,310 GBP per week
• 4,300 CAD per week

The fellowship lasts for four months full-time, with possible extensions.

Additional Benefits

Fellows also get:

• Direct guidance from Anthropic researchers.
• Access to shared workspaces in London, UK, and Berkeley, California.
• Options for remote work.
• Funding for computing

Frequently Asked Questions