Skip to main content
v2026.11,610 entries · CC-BY 4.0
LAC HealthLaboratory & ResearchLab & research supplies.Reagents, consumables, PPE & instruments — documented, fast, chain-of-custody shipping.Shop lac.us lac.us

Bezos Earth Fund: Climate Research Grants and the AI Grand Challenge

How the Bezos Earth Fund structures climate and nature research funding, and how its AI Grand Challenge for Climate and Nature program works for academic applicants.

The Bezos Earth Fund is Jeff Bezos’s climate- and nature-focused philanthropic research funder, established with a $10 billion commitment announced in 2020. It has since granted more than $2.4 billion across hundreds of awards to universities, NGOs, multilateral bodies, and research consortia. For academic research administrators, it is most relevant in two ways: as a large general climate-and-nature philanthropy funding applied and translational research, and, more specifically since 2025, as the funder behind the AI Grand Challenge for Climate and Nature — a dedicated program open to global academic institutions. This guide covers both, with emphasis on the Grand Challenge’s actual funding structure and eligibility terms. It is part of CASRAI’s series on private and philanthropic research funders — see also the IKEA Foundation and Rockefeller Foundation guides and the Grants Management & Research Funding pillar.

What the Bezos Earth Fund is

The Bezos Earth Fund is led by Executive Chairman Jeff Bezos and Vice Chairwoman Lauren Sánchez. It operates as a standalone philanthropic entity distinct from Amazon and from Bezos’s other ventures — a distinction research administrators should document the same way they would for any corporate-founder-affiliated foundation, since a grant from the Fund is philanthropic funding, not a commercial sponsorship arrangement with Amazon. The Fund organizes its work across seven program areas: Conserving & Restoring Nature, Future of Food, Environmental Justice, Decarbonizing Energy & Industry, Economics, Finance, & Markets, Next Technologies, and Monitoring, Data, & Accountability. Research funding is distributed across several of these areas rather than concentrated in one, which means a researcher evaluating fit should look at program-area alignment first rather than assuming a single “research grants” portal covers everything the Fund does.

Most Bezos Earth Fund grants go to established partner organizations — international NGOs, multilateral initiatives, and named research institutions — selected through the Fund’s own outreach and due diligence rather than through a standing open call. The AI Grand Challenge, covered below, is a notable exception: it is a competitive, application-based program that academic institutions can apply to directly.

The AI Grand Challenge for Climate and Nature

Launched in 2024 with a commitment of up to $100 million, the AI Grand Challenge for Climate and Nature is the Bezos Earth Fund’s dedicated program funding artificial-intelligence-driven approaches to climate and biodiversity problems. It is run through a program-specific site (aiforclimateandnature.org) separate from the Fund’s main grants pages, and it is structured in two funding phases rather than as a single award type.

Phase I: seed grants

In May 2025, the Fund announced 24 Phase I awards, each a $50,000 seed grant, totaling $1.2 million in initial funding. Phase I recipients included academic institutions alongside NGOs and startups — named university grantees reported at announcement included Cornell University, Delft University of Technology, King’s College London, the University of Leeds, and Wageningen University & Research. Phase I recipients also took part in an “Innovation Sprint” providing mentorship and team-building support alongside the funding itself, rather than a grant-and-forget disbursement.

Phase II: scale-up grants

Under the program’s published structure, up to 15 of the strongest Phase I projects are eligible for Phase II implementation grants of $2 million each, intended to fund development and scaling over a two-year period. Research administrators advising an investigator who has received Phase I funding should treat Phase II as a separate, competitive follow-on decision, not an automatic renewal — confirm current status and criteria directly with the program rather than assuming continuation.

Focus areas and eligibility

The Grand Challenge’s initial funding call centered on four areas: sustainable proteins, power grid optimization, biodiversity conservation, and open “wildcard” proposals for other climate-and-nature applications of AI. Phase I eligibility was open to both U.S.-based 501(c)(3) organizations and global academic institutions — a broader eligibility footprint than many U.S.-domestic philanthropic programs, which matters for research offices at non-U.S. universities evaluating whether they can apply at all.

Because this is a competitive, cohort-based program with specific application windows rather than a rolling grant mechanism, research administrators should verify current call status, deadlines, and any updated eligibility terms directly at aiforclimateandnature.org before an investigator invests proposal-development time — program structures and funding amounts for future cohorts are not guaranteed to match the Phase I terms described above.

Considerations for research administrators

  • Distinct from Amazon. Document Bezos Earth Fund awards as independent philanthropic grants in conflict-of-interest disclosure and funder-relationship records, not as Amazon corporate sponsorship, since the two are legally and operationally separate.
  • Two very different funding models under one name. The Fund’s general program-area grantmaking is largely by invitation to established partners; the AI Grand Challenge is a genuine open, competitive application process. Confirm which model applies before advising a PI on strategy.
  • Phased, milestone-linked funding. The AI Grand Challenge’s Phase I/Phase II structure means the initial award ($50,000) is intentionally small relative to the eventual scale-up grant ($2 million) — budget and staffing plans should reflect that Phase I is a seed/proof-of-concept stage, not the full project budget.
  • Verify current terms before proposal development. As with any philanthropic funder outside the federal system, award terms, reporting requirements, and indirect-cost treatment are set by the funder’s own grant agreement rather than a governmentwide regulation such as 2 CFR 200 — confirm these directly with the Bezos Earth Fund or aiforclimateandnature.org rather than assuming federal-grant conventions apply.

Frequently asked questions

Is the AI Grand Challenge for Climate and Nature a real, currently active program?

Yes. It is a genuine Bezos Earth Fund initiative, launched in 2024 with a stated commitment of up to $100 million, with its first cohort of 24 Phase I grants announced in May 2025. It is distinct from the Fund’s broader climate and nature grantmaking and is run through its own program site.

Can academic institutions outside the United States apply?

For the Phase I round, eligibility explicitly included global academic institutions alongside U.S.-based 501(c)(3) organizations, and the announced Phase I cohort included universities in the UK, the Netherlands, and elsewhere outside the US. Confirm current eligibility terms for any future round directly with the program, since terms can change between cohorts.

How much funding does a Phase I award actually provide?

The announced Phase I awards were $50,000 seed grants. Up to 15 of the strongest Phase I projects are eligible for a separate, competitive Phase II award of $2 million intended to fund two years of scale-up work — Phase II is not automatic and should not be budgeted as guaranteed.

Does the Bezos Earth Fund accept unsolicited research proposals outside the AI Grand Challenge?

Its broader program-area grantmaking is largely directed to established partner organizations through the Fund’s own outreach rather than an open unsolicited-proposal process. The AI Grand Challenge is the clearest example of a standing, application-based route for academic researchers; outside of it, direct outreach to the relevant program area is the more realistic path.

Referenced across the research world

University of Cambridge logoColumbia University logoCrossref logoUniversity of Edinburgh logoHarvard University logoUniversity of Oxford logoPrinceton University logoStanford School of Medicine logoUniversity College London logoORCID logoUniversity of Cambridge logoColumbia University logoCrossref logoUniversity of Edinburgh logoHarvard University logoUniversity of Oxford logoPrinceton University logoStanford School of Medicine logoUniversity College London logoORCID logo
  • University of Cambridge logo
  • Columbia University logo
  • Crossref logo
  • University of Edinburgh logo
  • Harvard University logo
  • University of Oxford logo
  • Princeton University logo
  • Stanford School of Medicine logo
  • University College London logo
  • ORCID logo

View CASRAI adoption →