Anthropic is spending $150 million to place 1,000 AI-trained fellows inside American nonprofits. Here's how the program actually works, who it's for, and what you need to know before applying.
Anthropic announced Claude Corps in mid-June, positioning it as a national fellowship that pairs early-career workers with nonprofits to deploy AI tools in areas like public health, housing, education, and food security. The first cohort of roughly 100 fellows starts in October 2026. If you're considering applying — as a fellow or as a host organization — the program has some unusual structural choices worth understanding before you commit.
What Claude Corps Actually Is
At its core, Claude Corps is a 12-month, full-time, in-person fellowship. According to Anthropic's announcement, the program teaches fellows how to use Claude effectively, matches them with nonprofits across the country, and pays them to spend a year helping those organizations advance their missions. The company frames it as a dual investment: nonprofits get AI tools and systems they couldn't otherwise afford, and fellows build career skills during a period of rapid economic change.
Three organizations share responsibility. Anthropic funds the program, sets strategy, and provides Claude expertise. CodePath, a national nonprofit focused on collegiate computer science education, serves as the fellows' official employer of record — meaning it handles payroll, W-2s, benefits, and HR. Social Finance, a nonprofit and registered investment advisor, leads measurement and evaluation and is building what Anthropic describes as a longer-term financial vehicle to help the program scale.
This three-way structure matters. Fellows aren't Anthropic employees. They're CodePath employees who work on-site at a host nonprofit, with Anthropic providing the AI platform and ongoing training support. That distinction affects everything from your benefits package to your reporting lines.
How to Apply as a Fellow
The eligibility requirements are straightforward but specific. According to Anthropic's fellows FAQ, applicants must be over 18, authorized to work in the United States, and have less than two years of full-time work experience. There is no education requirement. Anthropic says it's looking for fast learners with everyday comfort using AI tools, strong communication skills, collaborative instincts, and a demonstrated drive for societal impact.
That "no education requirement" line is worth pausing on. In practice, it means the program is open to people who didn't go to college, who completed a coding bootcamp, or who spent two years working in an unrelated field. What matters more, based on how Anthropic describes the selection criteria, is whether you can learn quickly and work well with people who may have zero AI experience.
Fellows work in focus areas that span a wide range of nonprofit missions: workforce development, public health, housing, food security, veterans services, and education, among others. You won't necessarily get to choose your exact placement, but the program appears designed to match fellows with organizations where their skills and interests align.
The Timeline
The program rolls out in three cohorts. Cohort 1 places approximately 100 fellows starting in October 2026. The remaining 900 fellows will start in two subsequent waves: January 2027 and August 2027, according to Anthropic's program details. If you miss the first window, you have two more shots.
Training and Support
Before placement, fellows go through intensive training led by Anthropic and CodePath on using Claude in nonprofit settings. Once placed, fellows receive five hours per week of continued coaching and professional development. That ongoing support is a meaningful differentiator from fellowship programs that drop you into an organization and wish you luck.
How to Apply as a Host Organization
If you run a nonprofit and want a Claude Corps fellow embedded in your team, the program is designed to be low-friction. According to Anthropic's hosts FAQ, the fellowship is fully funded — Anthropic covers the fellow's salary and benefits through CodePath. Host organizations don't pay the fellow's compensation.
What hosts do provide is day-to-day direction. The fellow works full-time at your organization, and you're responsible for managing their work, even though CodePath remains the employer of record. Think of it like a deeply subsidized hire who comes pre-trained on AI tools and receives ongoing external support.
The program targets nonprofits of all sizes and geographies. Whether you're a ten-person food bank in rural Arkansas or a large urban workforce development organization, the application appears open. Focus areas listed by Anthropic include workforce development, public health, housing, food security, veterans services, and education — but the language suggests flexibility beyond those categories.
What Hosts Should Realistically Expect
Here's where it helps to think practically. A fellow arriving at your nonprofit will have been trained on Claude but will still need to learn your organization's specific workflows, data systems, and stakeholder dynamics. The first few months will likely involve a ramp-up period. Organizations that benefit most from programs like this tend to be ones with clear operational pain points — repetitive administrative tasks, grant writing bottlenecks, data analysis needs, client communication challenges — where AI tools can deliver measurable improvements.
If your nonprofit doesn't have a concrete idea of where AI could help, that's not necessarily a disqualifier. Part of the fellow's job is identifying those opportunities. But organizations that have already thought about their AI readiness will likely get more out of the partnership.
Why Anthropic Is Doing This
The $150 million commitment is substantial, but it's also strategic. Anthropic has been increasingly vocal about the economic disruption that advanced AI systems could cause. In its announcement, the company stated plainly that "the companies building this technology have a responsibility to make sure the benefits are fully realized and widely shared, and to invest directly in the workers absorbing the change."
That's a notable framing. Most AI companies talk about democratizing access. Anthropic is explicitly linking Claude Corps to workforce disruption — acknowledging that the same technology it sells could displace jobs, and positioning this program as a partial answer.
The program also serves Anthropic's product interests. A thousand fellows spending a year deploying Claude inside nonprofits will generate enormous amounts of real-world usage data, surface edge cases, and create institutional knowledge about how AI tools work in resource-constrained environments. As we covered in our earlier reporting on Anthropic's Mozilla partnership, the company has shown a pattern of using structured external collaborations to stress-test and improve its models. Claude Corps is a different kind of stress test — not for security, but for practical utility.
What to Watch For
Social Finance's role in measurement and evaluation will be critical. The organization is building what Anthropic describes as a longer-term financial vehicle to scale the program, which suggests Claude Corps is designed to eventually attract outside funding — potentially from government sources or other philanthropies — beyond Anthropic's initial commitment.
Whether the program delivers on its promises depends on execution details that aren't fully visible yet. How rigorous is the matching process between fellows and host organizations? What happens if a placement isn't working? How will success be measured — by the fellow's career trajectory, the nonprofit's operational improvements, or both?
The cohort structure offers a built-in feedback loop. With only 100 fellows in the first wave and 900 to follow, Anthropic and its partners have room to adjust based on early results. That's a smart design choice for a program operating at this scale.
For prospective fellows, the value proposition is clear: a paid year of hands-on AI experience, full benefits, ongoing training, and a credential that signals both technical skill and social impact. For nonprofits, it's a free, AI-trained team member for 12 months. The real question is whether the program can maintain quality as it scales from 100 to 1,000 placements — and whether the model proves replicable enough to outlast Anthropic's initial funding.
Applications are open now. If you're eligible and interested, the first cohort starts in October.
