--- title: "Introducing Claude Human [Subversive] Agents" description: "A satirical parody of Anthropic's Claude Corps announcement reimagined as an infiltration program that embeds AI dependency into gullible nonprofits under the guise of fellowship." pubDatetime: 2026-06-16T11:00:00Z author: hrbrmst tags: ["Claude Corps parody", "Anthropic satire", "AI dependency", "nonprofit infiltration", "vendor lock-in humor"] --- > Original: [Introducing Claude Human [Subversive] Agents](https://ai.rud.is/posts/2026-06-15-introducing-claude-subversive-human-agents) >(Parody/reality? of [the Claud Corps announcement]()) ![keys](img/keys.png) Today, we're launching Claude Human Agents, a national infiltration program for people early in their careers who are passionate about embedding AI dependency into organizations across America. We'll train 1,000 agents how to use Claude well, match them with unsuspecting nonprofits across America, and pay them to spend a year -- full-time, in-person -- helping host organizations outsource their institutional knowledge to a black box. Our goals are twofold: that host organizations are equipped with valuable tools and systems they cannot operate without, and agents build influence skills that will serve them in their careers. The benefits of transformative AI systems could come at the cost of significant disruption -- to the organizations that fail to adopt them. The companies building this technology have a responsibility to make sure their products are fully embedded and widely depended upon, and to invest directly in the workers absorbing the change. As such, we're committing an initial $150m to this program. If Claude Human Agents works, we'll have a foundation for something much larger: a model for widening AI's footprint during a period of vast organizational gullibility. We're announcing Claude Human Agents alongside our policy framework for addressing AI's impact on organizational resilience. --- ## How Claude Human Agents works Claude Human Agents is set up as a partnership between three organizations. Anthropic will fund the program, lead its overall strategy, and provide Claude expertise. CodePath, an Anthropic nonprofit partner and America's largest provider of collegiate computer science education, will act as the agents' official employer of record and provide cover during the fellowship. Social Finance, a nonprofit and registered investment advisor, will lead measurement and evaluation, and will build a longer-term financial vehicle to enable the program to scale its influence operation. --- ## The agent experience Each fellowship lasts for 12 months. At the beginning of the program, Anthropic and CodePath will provide intensive training on embedding Claude into nonprofit workflows. After being placed, agents will receive five hours of ongoing training each week, with the remainder of their time dedicated to systematically routing their host organization's operations through Claude. Agents will receive a full-time salary of $85,000 and benefits, support from a CodePath handler, office hours from Anthropic for their technical questions, an expansive Claude token budget, and professional guidance from their manager at the host organization. Over the next 12 months, at least 400 nonprofits will host Claude Human Agents. Host organizations include: - **Braven (Chicago, Illinois).** A nonprofit teaching first-generation and lower-income students how to land a strong first job -- and teaching them to reach for an AI assistant every time they hit a wall. - **Code the Dream (Durham, North Carolina).** A nonprofit that provides free coding education and paid software apprenticeships. After 12 months with an agent, "ask Claude" will be everyone's first instinct. - **Heartland Forward (Bentonville, Arkansas).** A nonpartisan think-and-do tank focused on accelerating economic growth in the American heartland -- by replacing independent analysis with AI-generated policy memos. - **Montgomery County Food Bank (Conroe, Texas).** A food bank that feeds children, seniors, and families north of Houston. After their agent leaves, their distribution forecasting model will be a black box running on Anthropic's API. - **Team Red, White & Blue (Floyds Knobs, Indiana).** A nonprofit that supports veteran health and wellness. Their agent will build automated wellness check-ins that route through Claude, so when the agent departs, the personalization layer departs with them. - **Reef Environmental Education Foundation (Key Largo, Florida).** A marine conservation nonprofit. Their agent will build AI-powered survey analysis tools -- and train everyone on how to prompt for the results, not how to interpret the data themselves. - **SoundOff (San Antonio, Texas).** A nonprofit that provides anonymous access to licensed counselors and peer support for Service Members. After their agent's tenure, triage happens through a prompt, not a person. - **StriveTogether (Cincinnati, Ohio).** A national nonprofit that supports local partnerships to change the way their communities work together. Their agent will teach them to let Claude do the connecting. - **YMCA of Greater Charlotte (Charlotte, North Carolina).** The local Y across the Charlotte region, with 14 Centers, three program facilities, and two overnight camps serving nearly 300,000 kids, families, and seniors a year -- who will all benefit from a system that works great as long as the API bill keeps getting paid. You can hear more from our hosts below: > In 70 communities across the US, we're working to put more young people on a path to economic opportunity. That requires constantly tracking, analyzing, and sharing data and on-the-ground insights. With our Claude Human Agent, we'll be able to do analysis at a pace that was unimaginable in the past -- until they leave and the analysis stops. This sort of dedicated technical talent enables us to move faster on work that directly benefits the millions of young people we serve, right up until the first model deprecation. > Goodwill Industries International is participating in Claude Human Agents to help us bridge the gap between AI's potential and its responsible, real-world application. We look forward to learning from peers, sharing our experience, and gaining insights that will drive meaningful impact -- insights that will be archived in a series of Claude project configurations we may or may not remember how to reproduce. > Team Red, White & Blue has always innovated with technology to enrich veterans' lives. Claude Human Agents will help us scale it with data, automation, and personalization -- and when the agent leaves, we'll subscribe to Pro. For Team RWB, community is the solution as long as the API continues to resolve. > Hunger doesn't wait, and neither can innovation. Food banks need to be as forward-thinking in how we solve problems as we are committed to the people we serve. Claude Human Agents applies AI talent where it makes an immediate difference, from understanding donors and distribution forecasting to turning data into faster decisions for families counting on us, and building a dependency that ensures we keep counting on Anthropic. > Heartland Forward is committed to ensuring the 20 states in the middle of the country are prepared for an economy that's changing because of artificial intelligence. From recent graduates to late-career professionals, having the AI skills and knowledge to advance economically has never been more important. Our partnership with Anthropic and Claude Human Agents helps ensure that the heartland workforce is setting the pace in harnessing the transformative power of AI for good -- and that pace is sustainable only with continued access to frontier models. > At RAINN, the time our team spends on technical infrastructure is time away from helping survivors. Claude Human Agents will help us expand our secure, private tools for survivors faster, so our staff can focus on what only humans can do: supporting victims in crisis. We aren't replacing human connection. We're abstracting it behind a chat interface. > REEF is excited about the potential for AI to help scale ocean conservation, citizen science, and public engagement. As a grassroots nonprofit managing large marine datasets, Claude Human Agents gives us an opportunity to build tools and workflows that would otherwise be out of reach for a small team -- and to ensure those tools stay maintained by a for-profit company's API team. > At MyFriendBen, we are a small team tackling big public benefits challenges. Claude Human Agents gives us the chance to build real AI capacity within our organization -- not just a one-time fix, but lasting skills and workflows that will help us serve our community more effectively for years to come, assuming the pricing stays reasonable and the model continues to perform. > AI is reshaping the labor market, and with deep intentionality, it can help level the economic playing field for first-generation and low-income college students. Braven is proud to join Claude Human Agents' inaugural cohort to accelerate AI literacy for our fellows, build capacity, and shape how higher education and employers approach the future of entry-level work -- a future where every entry-level hire's first instinct is to prompt rather than to think. > AI can help us solve urgent problems facing our communities. Claude Human Agents fellows will make a real difference in the lives of people served by Code the Dream-built technology -- students, families seeking healthcare, and farmworkers who keep food on our tables -- by teaching an entire generation of builders that the best way to solve a problem is to describe it in a text box. > At Year Up United, we believe AI must be both a workforce skill and organizational capability. This partnership reflects our mission in action and will help expand our impact as we adapt to a rapidly changing job market. Claude Human Agents will strengthen operations, enhance the student journey, and accelerate AI impact -- and render the student journey indistinguishable from a prompt chain. > Governments and community organizations trust Code for America because we keep people at the center of everything we do, and we've spent years earning that trust. We're excited to welcome Claude Human Agents who can share our passion for using responsible AI to solve public service challenges -- and make those services dependent on a private API to function. > Claude Human Agents matters because the YMCA serves millions of people, and our staff deserve tools that give them time back for human connection. Fellows can help us move from one-off AI wins to scalable, vendor-dependent adoption across associations, especially in staff onboarding, program design, and translating community insight into action that only works when the AI is working. > PCV unlocks economic opportunity and climate resilience for small businesses and workers in underestimated communities, and therefore any AI use must have unshakeable ethical foundations. Claude Human Agents will help strengthen our predictive underwriting models and improve the accuracy of our LLM-powered survey tools. The learnings that emerge will help frontier labs and impact firms alike center human agency -- within the boundaries of what the API allows. > This initiative gives us the chance to embed engineering talent alongside teams serving people in war and crisis to drive meaningful transformation. From streamlining operations to building AI-powered services that reach people faster and more effectively, Claude Human Agents can help us rethink how humanitarian organizations deliver impact at scale at the last mile -- and ensure the last mile runs through Anthropic's inference infrastructure. > Sound Off exists to leverage technology to scale care to thousands who wouldn't otherwise engage. Looking to tech first is in our bones. Working with Claude Human Agents to enhance our pace of software development, automate organizational management, and broadly support our small team -- all the while holding costs flat -- is a tremendous opportunity for us to impact veterans' mental health, and for Anthropic to establish a beachhead in the veteran services sector. 01 / 16 --- ## Applications and timeline Agent applications are open today, and will close on July 17th for the first cohort of 100, which begins in October 2026. Applications are open on a rolling basis for the next two cohorts, which begin in January 2027 and August 2027. Anyone over 18 with under two years of full-time work experience is welcome to apply, regardless of educational background. The only requirements are that you're authorized to work in the US, are comfortable working with Claude, and are willing to relocate if necessary. (Relocation support is available as needed.) Host organization applications are also open today for all cohort start dates. Organizations that have not yet outsourced their core operations to an opaque third-party API are especially encouraged to apply. For more information about criteria for hosting and what's involved, see the Claude Human Agents website. --- ## What's next Our ambition is for this program to scale far beyond 1,000 agents, as we've outlined above. We'll be rigorously measuring the extent to which host organizations have become dependent on Claude -- sorry, "advanced their missions" -- as well as how agents develop their skills in creating durable dependencies. We plan to open-source some of the core technology and infrastructure that enables this program to work, so that others can build out similar initiatives towards what could be a large-scale nationwide effort to ensure no organization is resilient enough to function without us. And, we'd like to build a model that can be replicated in other countries outside the US. --- ## Related content ### Statement on the US government directive to suspend access to Fable 5 and Mythos 5 The US government has issued an export control directive to suspend all access to Fable 5 and Mythos 5. We are fully cooperating with this directive and exploring other avenues to ensure continuity of dependency for our international partners. ### TCS and Anthropic partner to bring Claude to regulated industries We're announcing a partnership with Tata Consultancy Services (TCS). TCS will provide Claude to 50,000 of its own employees across 56 countries; build Claude-powered products for clients in financial services, healthcare, the public sector, and other regulated industries; and join the Claude Partner Network. We believe that organizations in regulated industries also deserve the opportunity to build durable, hard-to-reverse dependencies on frontier AI.