Anthropic Launches 2026 Fellows Program: $15K+ Monthly Stipend, High-Impact AI Research, and No PhD Required

16th June 2026, Kathmandu

The race to build safe, reliable, and steerable artificial intelligence is accelerating, and Anthropic is opening its doors to the next generation of technical talent.

Anthropic Fellows Program

Applications are officially open for the Anthropic Fellows Program, a premium 4-month, fully funded research opportunity designed to fast-track technical minds into cutting-edge AI engineering and safety research.

Offering a massive stipend of approximately $15,400 per month, alongside extensive compute credits and direct mentorship from top-tier AI scientists, this program is a rare gateway into the frontier of AI development.

Whether you are looking to work from tech hubs like Berkeley or London, or want to contribute remotely from the US, UK, or Canada, here is everything you need to know about the upcoming September 2026 cohort.

Ready to apply? Skip the line and access the official application portal here: Apply for the Anthropic Fellows Program.

Program At A Glance: Compensation & Logistics

Anthropic views AI research as an empirical science. To allow fellows to focus entirely on solving massive technical puzzles, the program offers world-class compensation and infrastructure backing.

Because the weekly stipend is distributed based on your home currency, you can use the interactive breakdown tool below to see exactly how the financial tracking maps out over the 4-month research window.

Quick Logistics Overview:

Duration: 4 months of full-time research (40 hours/week), beginning late September 2026.

Locations: Dynamic shared workspaces in Berkeley, California, and London, UK, with fully remote-friendly options across the US, UK, and Canada.

Work Authorization: Candidates must independently hold valid full-time work authorization in the US, UK, or Canada. Note: Visa sponsorship is not available for this specific fellowship.

Breaking the Barriers: No PhD or Prior ML Experience Required

In a field often gatekept by advanced academic credentials, Anthropic is taking a radically inclusive approach. You do not need a PhD or extensive past publication records to apply.

Instead, the program targets raw technical capability, intense curiosity, and a deep alignment with Anthropic’s mission to build beneficial AI systems.

The primary objective for fellows is to utilize external infrastructure (open-source models and public APIs) to complete an empirical project that culminates in a public output such as a peer-reviewed paper.

Proven Track Record: Historically, over 80% of Anthropic Fellows successfully author and publish papers during their cohort, with over 40% of fellows transitioning into full-time roles at Anthropic.

Choose Your Track: The 5 Specialized Workstreams

To maximize impact, Anthropic has expanded the fellowship across five highly collaborative technical and policy workstreams. While you will default to being considered across multiple paths, you can indicate your preference during the application process:

1. AI Safety Fellows

Focused on reducing catastrophic risks associated with advanced AI systems. Research areas include Scalable Oversight (keeping superhuman models honest), Mechanistic Interpretability (deciphering the internal black box of LLMs), Model Organisms of Misalignment, and Adversarial Robustness.

Potential Mentors: Sam Bowman, Ethan Perez, Alex Tamkin, Trenton Bricken.

2. AI Security Fellows

Tailored for those with backgrounds in offensive security, vulnerability research, and pentesting. Past fellows’ projects include building scaffolds for control evaluations and utilizing autonomous AI agents to discover millions of dollars in blockchain smart contract exploits.

Potential Mentors: Nicholas Carlini, Keri Warr, Evyatar Ben Asher.

3. ML Systems & Performance Fellows

Ideal for strong software engineers who love building large-scale distributed systems and high-performance computing pipelines. Projects involve constructing CPU simulators for accelerator workloads, optimizing training data pipelines, and debugging model training processes.

Potential Mentors: Alwin Peng, Zygi Straznickas.

4. Reinforcement Learning (RL) Fellows

Centered around training, fine-tuning, and evaluating large language models. Fellows work directly on RL algorithms and build complex, safety-focused RL environments to enhance the safety and capabilities of flagship models.

Potential Mentors: Ruhua Jiang, David Brandfonbrener, Colt Steele.

5. Economics & Policy Fellows (The Anthropic Institute)

A dedicated track for interdisciplinary thinkers aiming to study AI’s economic footprint, labor market impacts, future workforce development, and the offense-defense balance of cyber capabilities.

Potential Mentors: Jack Clark, Marina Favaro, Maxim Massenkoff.

Are You a Good Fit?

While qualifications vary across workstreams, Anthropic is looking for candidates who meet these core technical baselines:

• Fluent in Python programming (Strict, non-negotiable requirement).

• Strong foundational backgrounds in Computer Science, Mathematics, Physics, Cybersecurity, or Economics.

• Execution speed: The ability to implement ideas quickly, think empirically, debug under uncertainty, and communicate complex concepts clearly.

• Mission-driven: A strong drive to permanently transition into full-time AI safety research.

A Note on Imposter Syndrome

Anthropic explicitly encourages applications from underrepresented groups in tech. If you are passionate about the social and ethical implications of AI but are hesitant because you don’t tick every single box, apply anyway. Diversity of perspective is core to their safety mission, and they urge candidates not to exclude themselves prematurely.

The Selection Process: What to Expect

The interview sequence for the Anthropic Fellows Program is calibrated tightly to industry research engineer standards. Because Anthropic reviews applications on a rolling basis, understanding the timeline layout is key to pacing your preparation:

How to Apply

Slots for the upcoming 2026 cohorts are highly competitive and fill up quickly due to the rolling review structure. Secure your spot at the frontier of AI evolution today.

For more: Anthropic Fellows Program