Ten Foundations, $18 Million, One Open Call: How Humanity AI Just Rewrote the Playbook for AI Public-Interest Funding

On May 12, 2026, ten of the most consequential names in American philanthropy made a joint announcement that, on its surface, looked like another in a season of headline AI grants. The Doris Duke Foundation, Ford Foundation, Lumina Foundation, Kapor Foundation, John D. and Catherine T. MacArthur Foundation, Mellon Foundation, Mozilla Foundation, Omidyar Network, David and Lucile Packard Foundation, and Siegel Family Endowment together unveiled Humanity AI, a pooled philanthropic vehicle dedicated to ensuring that artificial intelligence serves the public good. The initial tranche: $8 million to twelve inaugural grantees, $3 million to a flagship initiative called AI Civics, and a $10 million open call slated for summer 2026. Total commitment: more than $18 million.

But the real story is not the dollar figure. Eighteen million dollars, in the context of an AI industry that raised something like $300 billion in venture capital last quarter alone, is rounding error. The story is the architecture. Humanity AI is the first credible attempt by major foundations to pool their AI public-interest dollars, agree on a shared set of focus areas, write checks at a uniform size, and run a single coordinated open call. For the nonprofits, researchers, and community-led groups that have spent the past two years pitching ten different program officers in ten different vocabularies, that coordination is a small revolution.

This deep dive unpacks what Humanity AI actually funded, what the open call will probably look like, who is positioned to win, and where applicants should be focusing their thinking right now — not in summer, when the call opens.

What Was Actually Funded

Of the $11 million that has already been allocated, $8 million went to twelve inaugural grantees, each receiving $500,000. The uniform check size is itself a signal: this is general operating support, not project-specific funding, and the foundations want grantees treated as peers rather than a tiered hierarchy. The inaugural cohort:

• AI Now Institute — institutional AI policy research and accountability
• Center for Democracy and Technology — civil liberties in the digital era
• Center on Resilience & Digital Justice — community-driven AI repair
• Council on Foreign Relations LEAD AI — international AI governance
• Distributed AI Research (DAIR) Institute — community-rooted research independent of big tech
• Kinfolk Tech — kin-based and culturally grounded technology design
• Partnership on AI — multistakeholder AI safety and ethics
• Pulitzer Center — investigative journalism on AI's social impact
• Student Defense (SHAPE AI program) — protecting students from predatory AI in higher education
• TechEquity — labor and tech equity policy
• Data & Society — lead organization for the AI Civics initiative
• Digital Public Library of America — anchor partner on AI Civics

The separate $3 million AI Civics initiative, led by Data & Society with the Digital Public Library of America as anchor partner, is the most architecturally novel piece of the package. It is designed to embed community voice in how AI is created, deployed, and used, with public libraries serving as the convening infrastructure. That is a deliberate bet on libraries as the last neutral civic space — and a bet that the most credible AI public-interest work is downstream of community trust, not upstream of it.

The pattern across the cohort tells you what Humanity AI's funders actually believe. Five of the twelve grantees work on direct community impact (DAIR, Kinfolk, Center on Resilience, TechEquity, Student Defense). Three work on policy and governance infrastructure (CDT, AI Now, Partnership on AI). Two work on international and democratic guardrails (CFR LEAD AI, Pulitzer Center). And two anchor the AI Civics piece. That is a portfolio strategy. There is no token grant to a generic ethics lab and no enterprise-AI-friendly think tank. The cohort is built around the thesis that public-interest AI work has to live close to the communities being affected.

The $10 Million Open Call

This is where the strategic reading matters. Humanity AI has committed $10 million to a forthcoming open call, launching in summer 2026 with full criteria, timeline, and focus areas to be released "in coming months." Ford Foundation's Technology and Society director Lori McGlinchey framed the rationale: "As people's tolerance for unchecked technologies gives way to demand for greater agency, we must embed fundamental rights into design, deployment, and governance."

Three things to infer from the inaugural cohort and that framing:

First, expect community-led groups to be explicitly prioritized. Humanity AI's leadership has signaled an emphasis on "community-led groups closest to AI's impact." Translation: if your organization is a national policy think tank with no on-the-ground community arm, this is not your grant. If you are a regional labor advocacy group documenting algorithmic management abuses, you are precisely the target.

Second, expect the focus areas to mirror the inaugural cohort. The four named areas — safeguarding democratic institutions, protecting workers' rights, strengthening journalism, and advancing education — are not hypothetical. They are the lived portfolio. An applicant whose work fits cleanly into one of those four buckets, with clear theory of change and community accountability, is positioned. An applicant whose work fits three of them tangentially is not.

Third, expect a single shared application, not ten variants. The whole architectural point of a pooled fund is to spare grantees from running ten program-officer relationships in parallel. Applicants should prepare a single narrative tuned to Humanity AI's framing, not ten foundation-specific narratives.

Eligibility and Positioning

Humanity AI has not yet published formal eligibility criteria, but the inaugural cohort's composition is a strong signal. Eleven of the twelve grantees are 501(c)(3) nonprofits with established AI-and-society research or advocacy track records. One — Kinfolk Tech — represents the explicit inclusion of newer, community-rooted technical organizations. The implication for applicants: tax status alone is necessary but nowhere near sufficient. What the funders are buying is credibility on AI's social impact, demonstrated through publications, policy engagement, community partnerships, or technical work that has already moved a policy needle.

For organizations not yet at that level of demonstrated credibility, the open call is still worth tracking — but the strategic move is partnership. Pairing with an inaugural grantee or a similarly positioned organization is far more likely to land $500,000 than a standalone application from a new entity.

What Applicants Should Do Now

The open call has not opened. Criteria are not published. But three preparatory moves are worth making before summer:

1. Map your work to one of the four named areas with specificity. Generic "AI ethics" framing will not stand out. "Documenting algorithmic wage theft in the warehouse-logistics sector" will.
2. Document community accountability mechanisms. Who do you answer to? What governance structure ensures community input into your work? Foundations in this cohort have explicitly named community-led work as a priority.
3. Build a one-page theory of change keyed to the four focus areas. When the call opens, applicants who already have a sharp 250-word narrative will have a real time advantage over applicants drafting from scratch on deadline.

The Broader Significance

Humanity AI matters for reasons beyond its first $18 million. It is the most credible attempt to date by mainstream American philanthropy to coordinate AI public-interest funding at scale. If the open call goes well and grantee outcomes are documented over the next eighteen months, expect a second-round pooled fund — and expect other foundation collaboratives to copy the architecture. For applicants, that means Humanity AI is not a one-time opportunity but the beginning of a structural shift in how AI civil-society dollars get allocated. The organizations that win in 2026 are positioning themselves for a decade of coordinated philanthropic capital, not a single grant cycle.

The summer 2026 open call should be on the calendar of every AI-and-society nonprofit in the country. The work to be ready for it starts now.