Humanity AI

Humanity AI is a collaborative philanthropic initiative announced October 14, 2025, committing $500 million over five years to ensuring artificial intelligence serves the public good. Co-chaired by the Omidyar Network and the John D. and Catherine T. MacArthur Foundation, the founding coalition spans ten foundations: the Doris Duke Foundation, Ford Foundation, Kapor Foundation, Lumina Foundation, Mellon Foundation, Mozilla Foundation, David and Lucile Packard Foundation, and Siegel Family Endowment complete the partnership. Rockefeller Philanthropy Advisors serves as fiscal sponsor, managing the pooled grant fund and coordinating funder engagement through a dedicated Executive Director.

The initiative's animating claim, articulated by Omidyar Network's Michele L. Jawando, is that "AI is not destiny, it is design" — that the central question is not whether AI will transform society but whether people and communities outside Silicon Valley will have meaningful participation in determining how. The Humanity AI website frames the goal as building "a more human(e) future in which AI is shaped by and for people."

Five domains, three levers

Humanity AI's grantmaking and foundation-portfolio work run across five programmatic domains: Democracy (transparency, accountability, and civic participation in AI governance); Education (directing AI's benefits toward all students equitably); Humanities and Culture (protecting artists' intellectual-property rights while enhancing human creativity); Labor and Economy (workforce preparation and AI that enhances rather than replaces work); and Security (identifying and mitigating AI risks across critical sectors).

Three strategic levers drive the theory of change across those domains: expanding the diversity of decision-makers in AI design, development, deployment, and governance; strengthening organisations that amplify community voices in AI's future; and supporting public goods that democratise AI access and control. The operative funding standard is "community-led efforts" — work in which communities define goals and drive design.

Grantmaking: inaugural round and open call

Partner foundations began aligned grantmaking across their respective portfolios in fall 2025. The pooled fund's inaugural grant round, announced in May 2026, distributed more than $18 million in two streams.

A first stream of $8 million went to nine organisations at $500,000 each: the AI Now Institute, Center for Democracy and Technology, Council on Foreign Relations LEAD AI, the DAIR Institute, Partnership on AI, TechEquity, Kinfolk Tech, Pulitzer Center, and Student Defense. The announced focus areas for this stream — safeguarding democratic institutions, protecting workers' rights, strengthening journalism, and advancing education — map the initiative's five programmatic domains onto existing civil-society infrastructure. A $3 million grant established AI Civics, led by Data & Society with the Digital Public Library of America as anchor partner, to advance public education, literacy, and community decision-making on AI.

A second $10 million open call targets US-based nonprofits (deadline October 21, 2026) across Education, Labor and Economy, and Humanities/Arts/Culture. The call explicitly invites organisations "who have not typically seen themselves as having a role in shaping AI" — workers, artists, students, educators — and prioritises proposals using open-weight and open-source models where AI is central to the work, a structural preference that distinguishes this call from other AI philanthropy.

Position in the corpus funder slice

Against the corpus's funder admission test (2026-06-05 edge case: intersection of AI throughline AND small-d / grassroots engagement), Humanity AI passes both axes unambiguously. The AI throughline is structural and explicit — every programme and grant falls within the AI domain. The grassroots / civil society engagement axis is equally explicit: the initiative's design premise is that communities beyond tech hubs must co-determine AI's future, the open call targets workers and artists who have not previously engaged with AI governance, and the inaugural grantees include the corpus's own anchors for community-centred AI accountability research and practice — the AI Now Institute, the DAIR Institute, and Data & Society.

Within the corpus's funder slice, Humanity AI occupies a structural gap none of its ten founding members fills individually: the first coordinated pooled-fund vehicle channelling philanthropy specifically toward the civil-society layer of the AI-good movement. Five of the ten founding partners — Omidyar Network, the MacArthur Foundation, the Ford Foundation, the Kapor Foundation, and the Mozilla Foundation — are already in the corpus as individual funder entries; Humanity AI is the coordinating node through which their AI-good grantmaking flows collectively. The $500 million five-year commitment is the largest single AI civil-society philanthropic commitment in the corpus; the ten-foundation coalition is the largest coordinated funder partnership the corpus records.

funded_orgs is left empty per the schema's canonical-direction rule (Org ↔ Funder is canonically populated on the org side); the in-corpus grantee relationships with org-ai-now-institute, org-dair-institute, and org-data-and-society are recorded by adding fund-humanity-ai to those organisations' funders[] lists in this same session.