PitchBook Climate Tech Fundraising Data Shows a 40% Drop, While AI Revives Deals

PitchBook climate tech fundraising data shows specialist funds raised nearly 40% less capital in 2025, despite accelerating demand from artificial intelligence infrastructure. The decline creates an apparent contradiction. Climate-focused fund managers face a tighter market while energy, storage, cooling, and critical-material startups attract wider investor interest.

That contradiction is changing who finances climate technology and which companies receive attention. AI data centers need electricity, grid connections, cooling systems, backup power, and new construction materials. Those requirements have turned selected climate technologies into essential infrastructure rather than optional sustainability projects.

The shift does not mean the entire climate market has recovered. Capital remains concentrated in larger transactions and strategically important sectors. Specialist climate investors must also compete with generalist venture firms, corporate investors, infrastructure funds, and technology companies for the most attractive deals.

The central contest is therefore not climate technology against AI. It is specialist climate capital against broader infrastructure capital that follows AI demand. That distinction explains how climate-focused fundraising can fall while individual energy companies still secure unusually large investments.

PitchBook Climate Tech Fundraising Reveals a Split Market

The headline decline measures money raised by specialist investment firms, not every dollar invested in climate companies.

According to the original climate funding report, PitchBook found that fundraising by climate-specialist venture firms fell nearly 40% in 2025. The comparison was with 2024, rather than the sector’s earlier market peak.

Fund fundraising and startup financing describe different layers of the capital market. Fundraising measures commitments that limited partners make to venture firms. Startup financing measures the money those firms and other investors deploy into companies.

Those measures can move in opposite directions. A climate fund may struggle to raise its next pool of capital while an energy startup attracts generalist investors. Corporations, sovereign investors, infrastructure managers, and private equity firms can also support climate-related businesses without using a climate label.

This distinction matters because aggregate company funding looked much healthier than specialist VC fundraising. Silicon Valley Bank reported that United States climate technology companies received $29 billion in venture investment during 2025. That was the third-highest annual total in its dataset, behind 2021 and 2022.

The same climate investment data also showed how misleading a strong total can be. Ten large, late-stage transactions captured 28% of all investment. Deal activity slowed across most climate subsectors even while the headline dollar amount remained elevated.

Capital concentration creates two different realities. Companies connected to electricity generation, data center construction, industrial supply chains, or energy storage can attract substantial checks. Founders outside those preferred categories still face a selective and demanding financing environment.

The results also depend on how researchers define climate finance. BloombergNEF counted $77.3 billion in private and public climate-tech equity financing during 2025, an increase of 53% from 2024. However, that measure includes public offerings and other equity transactions beyond traditional venture rounds.

BloombergNEF found that venture funding for startups declined for a third consecutive year. Its equity finance analysis attributed much of the broader increase to public-market activity and major Asian transactions.

China led climate-tech equity funding with $24.5 billion, while the United States followed with $21.9 billion. Clean power, energy storage, and low-carbon transportation companies collectively raised $64.9 billion across public and private markets.

These figures are not contradictory once their scopes are separated. Specialist fund formation weakened. Venture deal counts slowed. Yet concentrated equity financing continued to reach businesses tied to urgent infrastructure needs.

That separation is the foundation of the current climate tech VC fundraising story. The market has not returned to broad enthusiasm. It has reorganized around projects that can address measurable industrial constraints.

AI Data Centers Are Rewriting Climate Investment Demand

AI has given selected climate technologies an immediate customer problem to solve, which is often more persuasive than a distant emissions target.

Training and serving AI models requires large data centers with substantial electricity needs. Developers must secure generation capacity, grid access, cooling, backup systems, and physical equipment. The resulting demand reaches far beyond chip suppliers.

Renewable power can be deployed comparatively quickly and at competitive costs in favorable regions. Its output varies with weather, however, so data center operators also need storage, transmission, demand management, or dependable generation.

Firm power refers to electricity that can be delivered when customers require it. Nuclear, geothermal, and certain long-duration storage systems can support that requirement. Many of those technologies remain expensive, slow to develop, or commercially immature.

PitchBook senior research analyst John MacDonagh described AI as both a headwind and a tailwind for climate technology. He judged the positive effect to be stronger because data center developers need new energy sources. He also cautioned that dependable generation technologies require significant development and commercialization capital.

That combination creates a broad investment map. Power generation addresses the data center’s core demand. Storage helps balance intermittent resources. Grid software manages constrained connections, while advanced cooling reduces electricity and water consumption.

Low-carbon steel and cement can reduce the construction footprint of new facilities. Alternatives to copper may relieve material constraints. Critical-mineral companies can support batteries, electrical equipment, and the wider industrial supply chain.

The commercial framing has also changed. A startup no longer needs to sell solely through a carbon-reduction promise. It can present its technology as a response to delayed grid connections, power scarcity, rising operating costs, or equipment shortages.

That shift expands the potential investor base. Generalist funds can evaluate revenue growth and customer demand without adopting a climate-specific mandate. Corporate investors can back technologies that protect their infrastructure plans.

AI is also competing for capital, talent, and management attention. An energy investment review based on International Energy Agency research found that AI investment increased among major energy investors. Energy specialization across those investors had slowed since 2022.

For the 50 large investors studied, AI’s share rose to 22% in 2025. Energy technology accounted for 16%, after reaching 17% in 2024. The underlying analysis covered investors that had deployed at least $1 billion overall and significant capital into energy innovation.

Those figures show why AI cannot be treated as an uncomplicated benefit. The sector creates new demand for energy technology while competing directly for venture allocations. It can strengthen the revenue case for climate startups without improving fundraising conditions for every climate manager.

The strongest companies sit at the intersection of both trends. They serve infrastructure buyers, address a physical bottleneck, and offer a plausible route from pilot deployment to repeatable sales.

Generalist Capital Is Challenging Climate Specialists

AI-driven climate deals increasingly attract investors that do not identify as climate investors, putting specialist firms under pressure to prove their distinct value.

Dawn Lippert, founder and CEO of Elemental Impact and a general partner at Earthshot Ventures, said generalist technology firms co-invest in 90% of Earthshot’s portfolio companies. Her figure illustrates how far climate financing has moved beyond specialist circles.

Generalist participation can increase the capital available to successful startups. It can also help companies reach software customers, industrial partners, and later-stage investors. Those relationships matter when a business requires several financing rounds before reaching commercial scale.

Climate specialists argue that they bring a different form of scrutiny. They examine environmental outcomes, community effects, industrial deployment barriers, and policy exposure. Those questions can reveal risks that a conventional growth investor might overlook.

However, specialist knowledge does not guarantee easier fundraising. Limited partners still compare climate funds with AI, enterprise software, defense technology, healthcare, and other strategies. They also consider liquidity, fund performance, interest rates, and the time required to return capital.

Hardware-heavy climate businesses often need more time and money than software companies. A successful laboratory result does not automatically create a scalable factory. Permitting, construction, customer qualification, and supply contracts can extend the path to revenue.

That timeline creates a structural challenge for traditional venture funds. A venture partnership usually operates within a fixed fund life. Some energy technologies require project financing or infrastructure capital after early technical validation.

Mike Schroepfer, Meta’s former chief technology officer, offers an example of the specialist response. His firm, Gigascale Capital, announced a climate-focused fund containing $250 million in June 2026. Schroepfer said he rejected suggestions that the vehicle should be renamed as an AI fund.

The decision preserved a clear climate mandate, but Gigascale still invests in a market transformed by AI. Schroepfer said energy, critical minerals, and supply-chain companies were seeing more investor capital than he had previously observed.

His experience captures the primary tension. Climate expertise remains useful, yet AI demand increasingly determines which opportunities appear commercially urgent. A specialist manager must participate in that demand without becoming indistinguishable from a general infrastructure investor.

Corporate participation adds another layer. Microsoft, Google, Amazon, and Meta joined Elemental Impact in a data center initiative announced in May 2026. The program uses operating facilities as potential test environments for climate technologies.

Elemental planned to invest between $500,000 and $5 million in as many as ten startups through 2027. The participating technology companies helped fund the program and pay membership fees. They did not formally commit direct investment capital to the selected startups.

The data center initiative targets cooling, energy storage, low-carbon materials, and related infrastructure. Its most important resource may be customer access rather than the investment amount.

Connecting a startup with a data center operator can shorten the route to a meaningful pilot. A successful deployment can generate performance data, customer references, and a clearer basis for further financing.

Yet corporate programs can also shape innovation around the immediate needs of a few large buyers. Technologies serving communities, smaller industrial customers, agriculture, or adaptation may receive less attention. Specialist investors must decide whether to follow concentrated demand or defend a wider climate mandate.

The AI Tailwind Does Not Reach Every Climate Startup

The renewed interest is narrow, capital-intensive, and tied to infrastructure constraints, so it should not be mistaken for a broad climate venture rebound.

The first limitation is concentration. Silicon Valley Bank’s $29 billion total for 2025 was boosted by a small group of large transactions. That pattern leaves many early-stage companies competing for fewer active investors.

A large financing round can dominate annual totals without improving seed or Series A conditions. It can also conceal falling deal counts. Founders should therefore examine stage, sector, and transaction volume rather than relying on one aggregate figure.

The second limitation is policy uncertainty. Silicon Valley Bank counted more than 50 federal actions since 2024 that created headwinds for climate technology. The bank cited weaker research capacity, funding reductions, permitting friction, and fewer tax incentives.

Policy changes can alter project economics even when customer demand remains strong. A company may need tax credits, federal procurement, loan guarantees, or faster approvals to move from demonstration to construction.

The third limitation is the cost of scaling physical technology. New reactors, grid equipment, energy storage systems, and industrial materials require factories, engineering teams, certification, and working capital. A convincing AI demand forecast does not remove those execution risks.

The fourth limitation comes from the customer itself. Data center developers are under pressure to bring capacity online quickly. They may prefer proven equipment when an experimental technology threatens construction schedules or reliability.

Startups must therefore show more than environmental benefits. They need evidence that their systems can meet uptime, safety, integration, and delivery requirements. Those standards can slow adoption even when a pilot receives enthusiastic support.

AI infrastructure can also increase fossil fuel use before cleaner supplies become available. New data centers may connect wherever generation and transmission capacity already exist. That can raise emissions and local electricity demand during the transition.

Communities are questioning water consumption, power prices, land use, and the distribution of economic benefits. Opposition can delay projects or force developers to change locations. Climate technologies attached to contested data centers inherit some of that political risk.

Microsoft’s chief sustainability officer, Melanie Nakagawa, acknowledged that communities are asking serious questions about energy, water, and local effects. She did not claim that the Elemental initiative would eliminate opposition.

The fifth limitation is measurement. Investors use different definitions for climate technology, energy technology, infrastructure, and AI-enabled climate solutions. A data center power company may appear in multiple categories, depending on the dataset.

That overlap complicates claims about recovery. BloombergNEF’s equity total includes financing types that PitchBook’s specialist-fund measure does not. Silicon Valley Bank’s United States venture data covers another distinct market boundary.

AI-enabled companies are especially difficult to classify. Software can optimize electrical loads, forecast generation, inspect industrial facilities, or model new materials. Yet attaching AI to a product does not automatically improve its climate impact or commercial durability.

Investors must test whether AI changes the underlying economics. Useful evidence includes lower operating costs, faster deployment, improved efficiency, reduced material use, or higher asset availability. A generic AI description provides no such proof.

Climate tech VC fundraising therefore remains under pressure even as selected deals accelerate. The AI tailwind rewards technologies linked to immediate capacity constraints. It does not repair the entire venture market.

What the Funding Numbers Say About Climate Technology

The market is rewarding climate companies that can translate physical scarcity into customer demand, while penalizing strategies dependent on broad thematic enthusiasm.

The strongest 2026 narrative is not a simple return of climate investing. It is a shift from climate as a category toward energy and industrial technology as operating requirements.

That change favors businesses with identifiable customers and measurable bottlenecks. A data center developer understands the value of obtaining more power or reducing cooling costs. An abstract promise about future emissions carries less immediate weight.

The distinction helps explain why investment totals can rise while specialist fund formation falls. Generalist capital can enter attractive transactions without funding the next generation of climate-focused managers.

It also explains why the climate label has become strategically complicated. Some founders avoid the term when customers or investors associate it with policy dependence. Others preserve it because climate expertise differentiates their technology and mission.

Neither choice changes the underlying economics. A startup still needs a product that works, a buyer willing to deploy it, and financing suited to its development timeline.

Specialist investors can remain relevant by solving problems that generalists often underestimate. These include industrial diligence, project finance, regulatory pathways, environmental accounting, and community engagement.

They can also connect early technical work with later pools of capital. A venture fund may finance product development, while an infrastructure investor funds deployment. Corporate customers can provide pilots, contracts, or equipment validation.

The risk is that specialists become scouts for larger pools without capturing enough of the resulting value. If generalist investors enter only after risks fall, specialist funds must secure ownership and follow-on capacity early.

Limited partners will watch those economics closely. A strong climate outcome does not automatically create a competitive venture return. Fund managers need exits, distributions, and evidence that their expertise improves investment selection.

Founders face a related choice. A company can pursue large data center customers, but dependence on a few buyers can weaken its negotiating position. Product design may also become tailored to one customer’s facilities.

Broader markets provide resilience. Grid operators, manufacturers, utilities, commercial buildings, and public agencies may need the same technologies. Companies that can serve several customer groups will have a stronger case than businesses dependent on uninterrupted AI construction.

Financial discipline is becoming more important across those groups. Silicon Valley Bank found that 52% of venture-backed climate companies reduced their net cash burn year over year. The bank linked that improvement to stronger gross margins and a focus on manufacturing efficiency.

Lower burn can extend a company’s runway when fundraising takes longer. It also provides evidence that management can scale without relying on repeated capital injections.

However, cost control cannot replace investment in testing, certification, or production. Climate founders must distinguish optional expansion from technical work required for commercialization.

The current market rewards that balance. Investors want exposure to AI-driven infrastructure demand, but they also want businesses that can survive if data center schedules slow. That is a more demanding standard than a thematic climate boom.

Three Signals Will Show Whether AI Can Sustain Climate Deals

The next phase depends on repeatable deployments, broader fund formation, and investment beyond a few unusually large transactions.

The first signal is conversion from data center pilots into commercial contracts. Programs run by Elemental Impact and large technology companies can introduce startups to buyers. The meaningful test arrives when those buyers order systems across multiple facilities.

Repeat deployments would strengthen the claim that AI creates durable climate demand. Isolated demonstrations would suggest that corporate programs provide visibility without establishing a scalable market.

Readers should watch which technologies move first. Cooling and energy-management systems may deploy faster than new generation technologies because they can fit within existing facilities. New reactors, geothermal projects, and long-duration storage require longer development cycles.

The second signal is specialist fund formation through the end of 2026. New closes would show that limited partners see climate expertise as valuable despite the popularity of broader AI strategies.

The size and number of funds will both matter. One large vehicle cannot establish a broad recovery. Emerging managers and sector-focused funds need access to capital if the market is to support diverse early-stage experimentation.

A continued decline would weaken the specialist model. It would indicate that climate companies increasingly depend on generalist, corporate, or infrastructure capital. That shift could narrow the technologies receiving early support.

The third signal is the distribution of startup financing. Investors should track deal counts, early-stage rounds, and the share of capital captured by the largest transactions.

Rising total investment with falling deal activity would confirm continued concentration. More seed and Series A financings across several subsectors would indicate a healthier pipeline.

Public and private financing must also remain separate in that assessment. A large public offering can improve an annual equity total without helping a young company finance its first commercial facility.

The same discipline applies to geography. China’s public-market transactions, United States venture rounds, and European industrial projects reflect different financing systems. Combining them can obscure the conditions founders actually face.

AI energy demand is real, but its investment effects remain uneven. It can attract capital toward generation, storage, cooling, materials, and grid technology. It can also absorb venture dollars, increase fossil generation, and intensify local opposition.

The PitchBook climate tech fundraising decline captures only one side of that market. The other side is a race to finance the physical systems required for AI growth.

For founders, buyers, and investors, the useful question is no longer whether AI supports climate technology in general. The question is whether a specific solution removes a costly constraint and earns repeat business.

Track contracts rather than pilot announcements, fund closes rather than stated interest, and deal distribution rather than headline totals. Those signals will show whether PitchBook climate tech fundraising conditions are stabilizing or whether AI is only creating a narrow group of winners.