AI Grants for Climate Tech: Funding the Models That Could Save the Planet

The Department of Energy announced $320 million in AI-for-science investments in December 2025. The Bezos Earth Fund has committed up to $100 million to an AI for Climate and Nature Grand Challenge. NSF is running a dedicated competition that pairs AI researchers with geoscientists and funds teams with up to $10 million over three years. The question for climate researchers and climate tech developers is no longer whether AI funding exists -- it is which programs match your work and whether your proposal is ready when the windows open.

For most of the last decade, climate science and AI funding occupied separate federal silos. Climate researchers applied to NOAA and DOE's Biological and Environmental Research program. AI researchers went to NSF's Computer Science directorate. That separation has collapsed, replaced by a fast-growing pool of interdisciplinary funding specifically designed for teams that can do both.

DOE's Genesis Mission and the Science of Foundation Models

The clearest signal in federal AI-climate funding came from DOE's Office of Science, which in December 2025 organized its investments under what it calls the Genesis Mission -- a coordinated push to build self-improving AI models trained on scientific data unique to the DOE's national laboratory system.

The $320 million flows through several vehicles. The Transformational AI Models Consortium (ModCon), under lab announcement LAB-25-3560, targets teams building foundational AI models that span climate, energy systems, and environmental science. The companion American Science Cloud provides infrastructure to host and distribute those models. Together they represent a bet that the next generation of climate models will be AI-native, not just AI-assisted.

For climate scientists, DOE's Biological and Environmental Research (BER) program is the most direct on-ramp. Starting in FY2026, BER shifted most funding to the Continuation of Solicitation for the Office of Science Financial Assistance Program -- a broad open-call rather than targeted notices. Climate and Earth system researchers should monitor this open call continuously. Program areas covering atmospheric process research, environmental systems modeling, and Earth-energy systems all now explicitly welcome AI integration.

NSF's CAIG Program and the Geoscience Bridge

NSF named the gap between AI researchers and Earth scientists and built a program specifically to close it. Collaborations in Artificial Intelligence and Geosciences (CAIG), solicitation number NSF 25-530, funds interdisciplinary teams of two to three senior researchers, with NSF estimating between five and nine awards per competition totaling $6 million to $10 million per cycle.

The program runs on annual competition cycles. The 2025 full proposal deadline was April 2, 2025; the 2026 competition has a full proposal deadline of February 4, 2026. CAIG proposals are not meant to be AI-showcase projects that mention climate as an afterthought. NSF is explicit that the science question must drive the work, with AI serving as the enabling method to answer questions that were previously intractable. Teams without genuine geoscience collaborators have consistently fared poorly in review. Research areas that have performed well include hurricane track prediction beyond traditional numerical weather prediction skill, seasonal drought forecasting, and ice sheet dynamics modeling using sparse observational data.

The Bezos Earth Fund's $100 Million AI Challenge

Outside the federal system, the largest AI-climate funding program in the philanthropic space belongs to the Bezos Earth Fund, which launched its AI for Climate and Nature Grand Challenge with a commitment of up to $100 million.

The program has already moved through two phases. In May 2025, the Fund awarded 24 Phase I grants of $50,000 each, totaling $1.2 million in initial feasibility awards. In October 2025, 15 of those teams advanced to Phase II, each receiving up to $2 million over two years to scale real-world applications. Phase II awardees ranged from teams using AI to decode threatened bird vocalizations to groups automating plant species identification for conservation monitoring.

The challenge focuses on three domains: sustainable food systems and alternative proteins, power grid optimization for clean energy, and biodiversity monitoring. The Fund has also reserved space for wildcard solutions that do not fit those categories but demonstrate transformational potential.

No Phase III timeline has been announced, but the $100 million commitment provides room for future rounds. Organizations working on AI for ecological monitoring or food systems decarbonization should watch the Grand Challenge site for future calls.

Climate Change AI's Innovation Grants -- and a NOAA Warning

For research teams that sit outside the major federal or large-foundation tracks, Climate Change AI runs an Innovation Grants program that has become one of the most accessible entry points in the field. Awards reach up to $150,000 for one-year projects, with an emphasis on publicly released datasets, open-source tools, and reproducible research. The program funds mitigation, adaptation, and Earth system science -- and actively encourages teams from low- and middle-income countries, reflecting the global distribution of climate risk. The 2024 cycle closed September 15, 2024; the program has run annually since 2022, suggesting another cycle in 2025. Strong proposals tend to tackle narrow, well-scoped problems rather than broad "AI for climate" ambitions without a defined scientific contribution.

The contrast with NOAA is instructive. NOAA has deployed real AI weather infrastructure -- its AIGFS and AIGEFS models run operationally and extend forecast skill by 18 to 24 hours over traditional methods -- but the proposed FY2026 budget calls for eliminating the Office of Oceanic and Atmospheric Research entirely, cutting more than $50 million from climate research grants and labs. Congress has not enacted this as of early 2026, and bipartisan legislation has been introduced directing NOAA to accelerate AI forecasting for wildfires and extreme weather. But the uncertainty is real enough that researchers who have historically relied on NOAA grants should be diversifying toward NSF, DOE, and the philanthropic programs above.

Building a Proposal for This Moment

The distinguishing characteristic across every major AI-climate program is the insistence on genuine interdisciplinarity. DOE's Genesis Mission wants AI teams that understand atmospheric chemistry. NSF CAIG wants geoscientists who can collaborate with machine learning researchers. The Bezos Earth Fund advanced teams with documented field partnerships, not teams that had built impressive models in isolation from ecological reality.

Proposals that lead with the AI technology -- "we have a foundation model and want to apply it to climate" -- consistently underperform against proposals that lead with the scientific or operational problem and then explain why AI is the right tool. That framing distinction matters at every stage from reviewer scoring to program officer selection.

Data provenance is the other recurring differentiator. Every major funder wants to know what training data your AI system will use, how its quality has been validated, and what you will contribute back to the scientific community. Proprietary data strategies are a competitive disadvantage in nearly every AI-climate track.

Microsoft's AI for Earth program provides a supplementary path for teams that need compute infrastructure more than direct cash, accepting rolling applications across biodiversity, climate, agriculture, and water. The company's Climate Innovation Fund has separately committed $1 billion to climate technology with over $800 million already deployed -- a signal of where corporate R&D is flowing even as federal budgets remain contested.

For researchers ready to move from watching this landscape to competing in it, Granted can help match your team's specific expertise to open solicitations and move you from rough concept to submission-ready draft before the next deadlines close.

Sources:

• DOE Energy Department Advances Investments in AI for Science
• DOE Office of Science Open FOAs
• NSF 25-530 CAIG Solicitation
• Bezos Earth Fund AI for Climate and Nature
• Bezos Earth Fund $30M Phase II Announcement
• Climate Change AI Innovation Grants
• NOAA AI Center
• NOAA AI-driven Weather Models
• Microsoft AI for Earth Grants
• Microsoft Climate Innovation Fund