10 of the Biggest U.S. Foundations Just Pooled $18 Million for Public-Interest AI. The $10 Million Open Call Is Coming This Summer.

Ten of the most consequential private foundations in the United States — Ford, Mellon, the John D. and Catherine T. MacArthur Foundation, Doris Duke, Lumina, Kapor, Mozilla, Omidyar Network, the David and Lucile Packard Foundation, and the Siegel Family Endowment — just put their names on the same grant announcement. That alone is news. The pooled vehicle they created, Humanity AI, is now publicly funded at more than $18 million, with $8 million already awarded to 12 inaugural grantees and $10 million reserved for an open call this summer.

For organizations working at the intersection of AI and the public interest, this is the foundation event of 2026. Pooled funds of this scale and this configuration are uncommon. When they do happen — the Audacious Project, the Global Innovation Fund, the Co-Impact collaborative — they reshape how the field thinks about scale, coordination, and what "credible" looks like to foundation program officers. Humanity AI is going to do the same thing for public-interest AI work.

Foundation grants moving fast in your field? Browse our Foundation Grants Hub for a continuously updated database of private foundation opportunities across AI, technology, journalism, education, democracy, and economic equity.

Why a Pooled Fund, and Why Now

Pooled philanthropic funds emerge when the funders themselves believe no single foundation can move the needle alone. The pattern is consistent: a topic that is bigger than any one program officer's portfolio, an urgency that does not match the typical foundation timeline, and a recognition that grantees were being whipsawed by inconsistent funder priorities across the ecosystem.

That is the diagnosis Humanity AI's founding partners appear to share about public-interest AI. Generative AI has moved from a niche concern to a mainstream policy and labor issue inside three years. Foundation portfolios that previously funded "responsible technology" or "internet freedom" suddenly have to decide whether to absorb AI work into existing programs or stand up dedicated AI strategies. Most chose the latter, and most discovered they were duplicating reviews, asking grantees the same questions, and funding the same handful of well-positioned organizations.

A pooled fund solves three problems at once. Foundations can run one due diligence process rather than ten. Grantees can write one proposal rather than ten. And the field as a whole can make commitments large enough to actually move strategy at the grantee organizations, rather than fragmenting a $500,000 grant into five $100,000 checks that each require their own reporting.

What is striking about Humanity AI specifically is the breadth of the founding partners. Ford, Mellon, and MacArthur are legacy general-purpose foundations. Lumina is an education-only funder. Kapor, Mozilla, and Omidyar have explicit technology missions. Doris Duke focuses on the arts, the environment, child well-being, and Islamic studies. Packard is biodiversity-and-population. Siegel has a workforce-and-society lens. The fact that this group converged on a single AI vehicle suggests every one of these foundations now sees AI as cross-cutting through its existing mission — not as an adjacent technology issue.

What the 12 Inaugural Grantees Tell You

The $8 million inaugural cohort is 11 organizations receiving $500,000 each, plus a single $3 million commitment to Data & Society for its AI Civics initiative, with the Digital Public Library of America named as the community engagement partner. The cohort:

• AI Now Institute — labor, climate, and government accountability research on AI
• Center for Democracy and Technology — digital civil rights and liberties advocacy
• Center on Resilience & Digital Justice — accountability and repair for AI harms
• Council on Foreign Relations LEAD AI — policy analysis and public discourse
• DAIR Institute — community-driven AI research and grassroots knowledge-building
• Kinfolk Tech — art, technology, and collective memory initiatives
• Partnership on AI — multi-stakeholder collaboration and research
• Pulitzer Center — global journalism training on AI reporting
• Student Defense / SHAPE AI — institutional guidance on responsible AI adoption
• TechEquity — technology-industry accountability for economic equity
• Data & Society + DPLA — AI Civics initiative

Read the list with the four declared focus areas in mind: democratic institutions, workers' rights, journalism, and education. Every grantee maps cleanly to at least one focus area, and several map to two. This is not a portfolio assembled by sampling resumes — it is a deliberate distribution that ensures Humanity AI can credibly claim coverage of each focus area on day one.

Three patterns are worth extracting if you are positioning for the summer open call.

The pooled fund is buying capacity, not projects. Most of the inaugural awards are unrestricted or lightly restricted general operating support sized to fund a meaningful slice of an organization's annual budget for one to two years. The fund is not buying a research paper, a pilot program, or a conference. It is buying the institutional capacity of organizations the founders already knew and respected. That has implications for the open call: applicants who frame their proposals as projects will lose to applicants who articulate organizational strategy.

Community-led organizations have a real seat. DAIR Institute, the Center on Resilience & Digital Justice, and TechEquity are explicitly community-driven or community-rooted. The published framing of the open call emphasizes "supporting organizations led by communities most affected by AI." That is consistent with what the inaugural cohort actually demonstrates rather than aspirational language — it is the operating commitment.

The cohort is research, advocacy, journalism, and arts in one fund. That is unusual. Most foundation portfolios separate these by program area. Humanity AI's bet is that public-interest AI requires research feeding advocacy feeding policy feeding journalism feeding cultural narrative — and that funding the pieces in isolation has failed for fifteen years on internet policy. Applicants for the summer call should think about where they sit in this ecosystem and articulate it explicitly. "We are a research organization, but our work feeds policy advocates we have named, and we have a journalism partner that translates our work for public audiences" is a stronger frame than "we are the best research organization in the field."

Positioning for the Summer Open Call

The $10 million open call is forthcoming this summer. Specifics on application timeline, eligibility, and grant sizes have not been published yet. But the contours are predictable from the inaugural cohort, the published framing, and the founders' individual histories.

Expected grant sizes. Based on the inaugural pattern, expect grants in the $250,000 to $1 million range, with multi-year commitments. The $3 million Data & Society commitment is an outlier representing the AI Civics anchor initiative; most open-call awards will be smaller.

Likely eligibility. Expect 501(c)(3) public charities and fiscally sponsored projects to be the primary eligible category. Some pooled funds also accept 501(c)(4) advocacy organizations or international NGOs through intermediaries; Humanity AI has not yet clarified its position, but the journalism, advocacy, and policy work funded inaugurally suggests latitude.

Likely review criteria. Pulling from the founders' shared language: alignment with the four focus areas, demonstrated community leadership or accountability mechanisms, evidence that the organization can absorb and deploy multi-year funding effectively, and a plausible theory of how the work changes outcomes for affected communities rather than merely producing knowledge.

What to start now. Three concrete preparation steps for organizations considering an open-call submission.

First, build the case for organizational capacity, not project capacity. The inaugural cohort suggests reviewers will be looking at audited financials, board governance, staff retention, and whether the organization has demonstrated the ability to operate at the scale Humanity AI is funding. If your organization has never received a $500,000 single-funder commitment, your case for absorbing one needs to be explicit.

Second, articulate your community accountability mechanism. The founders' emphasis on "communities most affected by AI" is not rhetorical. Organizations that can describe in concrete terms — a community advisory board with decision rights, governance representation, a participatory grantmaking subprogram — will outscore organizations that gesture at community-centered values without operational evidence.

Third, identify ecosystem partners. The inaugural cohort is densely networked; Data & Society and DPLA are explicitly partnered, and most of the other grantees have public collaborations with at least two others. If you are planning to compete on your own, you are competing against organizations that walked in with ecosystem letters already signed.

What This Means for the Rest of Foundation Philanthropy

Humanity AI is large enough to set the floor for what other foundations consider serious AI commitments. Family foundations and corporate philanthropies that have been writing $25,000 checks to AI organizations will start hearing the question "have you applied to Humanity AI?" as a credibility test for whether the grantee is operating at a serious scale. That is a double-edged outcome for the field. Organizations that fit the Humanity AI profile will see capital flow more easily. Organizations that do not — smaller community-rooted groups, single-issue advocacy organizations, regional efforts — will need to articulate their non-Humanity AI niche more carefully.

For grant strategists at organizations of any size, the action this quarter is the same. Read the four focus areas carefully. Look at the inaugural cohort and identify the gap your organization fills. Get your application infrastructure — capacity narrative, financials, community accountability documentation, ecosystem partnerships — ready before the open call drops. When the summer announcement lands, the difference between a competitive submission and a non-competitive one will be measured in weeks of preparation that should already be underway.

Ten foundations rarely agree on anything. They have agreed on this. The summer open call is the highest-leverage foundation moment of 2026 for organizations working on public-interest AI. The work to be ready for it starts now.