Anthropic Is Hiring 1,000+ People in AI — Here's the One Thing Every Application Wants

What Are the Two Anthropic Opportunities — and Who Can Apply?

Anthropic is running two very different programs at the same time, and together they're opening more than a thousand paid positions in AI.

The first is the Anthropic Fellows Program, a research fellowship that pays roughly $3,850 a week for about four months, plus benefits and compute funding. Fellows work alongside Anthropic researchers across tracks like AI Safety, AI Security, ML Systems, Reinforcement Learning, and Economics & Policy. Some tracks lean technical; others — especially Economics & Policy — explicitly welcome non-traditional backgrounds.

The second is Claude Corps, announced June 11, 2026. It's a $150 million program that hires 1,000 early-career people, pays them $85,000 a year plus benefits, and places them inside nonprofits for 12 months to help those organizations actually use AI. The first cohort of 100 starts in October 2026, with applications closing July 17, 2026 — less than a month away. The remaining fellows join later cohorts in January and August 2027.

The eligibility bar is refreshingly low on paper: you must be 18 or older, have U.S. work authorization, fewer than two years of full-time work experience, and be comfortable working with Claude. No college degree required.

Apply or learn more:

• Claude Corps: anthropic.com/news/claude-corps
• Anthropic Fellows: job-boards.greenhouse.io/anthropic

What Do These Applications Actually Want From You?

Here's the part most people miss. "No degree, no experience" gets the clicks — but it doesn't mean no preparation. When you read the fine print, a consistent theme shows up: these programs want people who have already built things.

One detailed program analysis put it plainly — Claude Corps is aimed at people who have already built independent projects, automations, or tools. In other words, they're not looking for a diploma. They're looking for proof of building.

That proof usually lives in one place: a public portfolio of your work. That can look like a GitHub profile with project repositories, a personal website or portfolio, or links to things you've actually made and can explain. A repository — "repo" for short — is simply a project folder that holds your code and its history; your GitHub account can hold many of them. Each one is a small piece of evidence that you don't just talk about AI, you build with it.

You don't need to be a senior engineer. But you should be able to point to something real, and explain at a basic level what it does and how it works. If your project happens to include some Python — which most do under the hood — even better. You don't need to announce "I know Python." The project quietly proves it for you.

How Do You Actually Build Something Worth Showing?

The best way to build a portfolio isn't to invent a fake "demo" project. It's to build something you would actually use — and then keep using and updating it. Real use forces you to fix real problems, which is exactly the experience these programs (and employers) want to see.

Start small. Pick a problem in your own life or work, then use Claude to build a simple tool that solves it. Here are three starter prompts you can paste into Claude to get moving:

Prompt 1 — Pick a project:

"I want to build my first real coding project to show I can build with AI. I have no coding background. Ask me about my daily routine, work, and frustrations, then suggest 3 simple but genuinely useful tools I could build and actually use — each with a one-line description of what it does."

Prompt 2 — Build it (with Python):

"Let's build [the tool I picked]. Use Python. Walk me through it step by step, write the code, and after each section, explain in plain English what that part does so I can understand and explain it myself later."

Prompt 3 — Make it public and explainable:

"Help me put this project on GitHub. Write a clear README that explains what the tool does, how to run it, and what I learned building it — written so a non-technical person could understand it."

That third step matters more than people realize. Being able to explain your project — what it does, why you built it, what the code is roughly doing — is often what separates a strong applicant from a forgettable one.

Turning a Project Into Income — Not Just an Application

Here's where it gets interesting. The same skill that gets you hired by an AI company is the skill that lets you build something people will pay for. A tool you build for yourself can become a tool you sell to others. That's the leap from "I built a project" to "I built a product."

That's exactly what I help people do inside my SaaS Launch Sprint — a done-with-you program where we take your idea and build and launch a real SaaS platform using Claude, even if you've never coded. You walk away with something live, something you can show, and something you can actually monetize. Whether your goal is landing one of these Anthropic roles or building your own income, the foundation is the same: stop consuming, start shipping.

👉🏾 Learn more about the SaaS Launch Sprint

But to be clear — you do not need to buy anything to start. Use the prompts above. Build the thing. Make it public. That alone puts you ahead of most applicants.