Climate tech VC funding hit $26.1 billion in H1 2026, up 55% year on year — but a lot of that is low-carbon data-center infrastructure (roughly a third of the half, including two mega-deals), not pure climate applications. The interesting signal is elsewhere: a smaller, later-stage group of companies where AI genuinely is the product — satellite pattern recognition, real-time grid orchestration, autonomous building control — not a label stapled onto a climate pitch deck.
We covered whether AI can actually solve global warming in depth already; the honest answer is that it can't by itself, but it measurably helps in specific, verifiable places. This list is those places, in company form: 10 AI and climate tech startups actually shipping in 2026, what each one builds, who funds them, and how central AI really is to each — including where it's mostly deep-tech engineering with AI as one input, not the whole story.
Quick reference: the 10 startups
| # | Startup | What it builds | AI's role | 2026 signal |
|---|---|---|---|---|
| 1 | Watershed | Enterprise carbon accounting | Core — automated data extraction, categorization | $1.8B valuation, $100M Series C |
| 2 | Emerald AI | Grid-flexible AI data centers | Core — real-time workload/power orchestration | $25M round, Nvidia-backed, 16-month total $68M |
| 3 | Kayrros | Satellite emissions intelligence | Core — ML over Copernicus satellite imagery | Acquired by Energy Aspects, May 2026 |
| 4 | GHGSat | Satellite methane/CO2 quantification | Core — detection + quantification models | Constellation-based, facility-level attribution |
| 5 | Overstory | Vegetation-risk monitoring for grids | Core — ML over satellite + climate data | Wildfire/outage risk for utilities |
| 6 | Fervo Energy | Enhanced geothermal drilling | Supporting — subsurface modeling | 2026 IPO, $1.9B raised, shares +35% |
| 7 | Pachama / Carbon Direct | Forest carbon MRV | Core — satellite + AI verification | Acquired by Carbon Direct, Nov 2025 |
| 8 | ClimateAi | Agricultural climate-risk forecasting | Core — hyper-local weather/yield models | $38M raised, 30+ enterprise customers |
| 9 | PassiveLogic | Autonomous building infrastructure | Core — generative AI control ("physical AI") | $74M Series C, $125M total, Nvidia-backed |
| 10 | AirMyne | Direct air capture | Supporting — process/thermal optimization | ENEOS partnership, CA sequestration site in 2026 |
How this list was built
Every company here had to clear two bars: it's shipping something measurable in 2026 (not a pilot deck), and AI plays a real, checkable role in the product — not a rebrand of an existing climate-tech category. Funding, customer, and status figures are pulled from primary company sources, press releases, and 2026 reporting, cited inline. Where a company's AI role is secondary to its core engineering (geothermal drilling, direct air capture chemistry), that's called out rather than glossed over — see our evidence-based assessment of what AI can and can't do for climate change for the same discipline applied across the sector.
1. Watershed — enterprise carbon accounting
Watershed markets itself as "the sustainability AI platform," and the AI is doing real work — automating supplier-level emissions data extraction, categorization, and reporting across a customer's supply chain, instead of manual spreadsheet accounting. It raised a $100 million Series C valuing the company at $1.8 billion, backed by Sequoia, Kleiner Perkins, and Emerson Collective, with customers including Airbnb, Walmart, Stripe, BlackRock, Spotify, and four of the top six US banks.
Why it matters: carbon accounting is the measurement layer every other climate claim depends on — including the ones later in this list. Verdantix named Watershed a leader in its 2026 Enterprise Carbon Management Software report.
2. Emerald AI — grid-flexible AI data centers
Emerald AI solves a problem this blog has covered from the data-center energy angle before: AI infrastructure growth is outpacing grid capacity. Its software orchestrates AI compute workloads in real time so data centers can flex power consumption up and down, turning them into grid-supportive assets instead of pure load.
Nvidia's NVentures backs it, alongside Eaton, GE Vernova, Radical Ventures, Salesforce, and Samsung — an $25 million strategic round in March 2026 brought total funding to $68 million in 16 months. It's working with Digital Realty, EPRI, and PJM Interconnection on a large-scale power-flexible facility in Virginia in 2026.
Why it matters: this is one of the few startups turning AI's own power demand into part of the grid solution, not just the problem.
3. Kayrros — satellite emissions intelligence
Kayrros processes millions of satellite images daily from the European Space Agency's Copernicus constellation to track methane emissions globally, with AI doing the pattern-detection and attribution work at a scale no analyst team could match manually. It was acquired by Energy Aspects in May 2026 and launched a Nature Impact Platform (biodiversity intelligence) in March 2026, extending the same satellite-plus-AI approach beyond methane.
Why it matters: methane has an outsized near-term warming effect, and detection-to-repair speed is the actual lever — Kayrros's AI is the detection half of that pipeline.
4. GHGSat — satellite methane and CO2 quantification
GHGSat operates its own satellite constellation purpose-built to detect and quantify methane and CO2 emissions from individual industrial facilities — not just regional estimates, but facility-level attribution precise enough for operators and regulators to act on.
Why it matters: facility-level attribution is what turns a detection into an accountable, fixable leak, which is the gap between "we found emissions" and "someone actually repaired them."
5. Overstory — vegetation-risk monitoring for utility grids
Overstory uses machine learning over satellite imagery and climate data to model where vegetation growth threatens power lines — a direct input to wildfire risk and outage prevention for utilities managing thousands of miles of grid infrastructure they can't physically walk every season.
Why it matters: utility-caused wildfires are one of the more direct, physical-world harms climate change amplifies, and this is a case where a better prediction genuinely changes a maintenance decision.
6. Fervo Energy — AI-optimized enhanced geothermal drilling
Fervo Energy had one of 2026's biggest climate-tech exits: an IPO that raised $1.9 billion, with shares rising 35% on debut. Its core technology is enhanced geothermal systems — using techniques borrowed from oil-and-gas horizontal drilling to access geothermal heat anywhere, not just at existing hot spots — with subsurface modeling assisted by machine learning.
Worth being precise here: geothermal drilling is fundamentally a deep-tech engineering achievement, and AI is a supporting tool for subsurface prediction, not the headline technology. It earns a spot on this list for scale and market validation, not for being an "AI-first" company.
Why it matters: geothermal is one of the few renewable sources that can deliver baseload power, and Fervo's IPO is a signal institutional capital sees this as investable at scale now.
7. Pachama (now part of Carbon Direct) — forest carbon verification
Pachama built AI-enabled tools to monitor, report, and verify forest-carbon projects using satellite data — addressing the long-standing credibility problem in forest carbon credits, where claimed sequestration often didn't match reality on the ground. It was acquired by Carbon Direct in November 2025, and its MRV technology now operates under Carbon Direct rather than as an independent brand.
Why it matters: carbon-credit integrity has been one of the weakest links in climate finance, and AI-based satellite verification is a direct answer to "how do we know this forest is actually still standing."
8. ClimateAi — agricultural climate-risk forecasting
ClimateAi runs a hyper-local weather and climate-risk platform (ClimateLens) for agricultural and water supply chains, serving 30+ enterprise and government customers across 35 countries. Its recent ClimateLens Monitor Yield Outlook offers AI-driven crop-yield forecasts for commodities including corn, wheat, soybeans, and canola.
Why it matters: agricultural supply chains are among the most climate-exposed parts of the economy, and better yield forecasting has a direct line to food-security and pricing decisions, not just an interesting dashboard.
9. PassiveLogic — autonomous building infrastructure
PassiveLogic's Hive platform applies generative AI to real-time autonomous control of physical infrastructure — hospitals, office towers, data centers — what the company calls "physical AI." It raised a $74 million Series C (total funding over $125 million) with Nvidia's NVentures and Johnson Controls among the backers, and projects the autonomous-buildings market could reach $1.3 trillion by 2030.
Why it matters: buildings and industrial control are one of the most defensible near-term AI climate applications because energy meters give a direct, measurable outcome — PassiveLogic is betting the entire building-management layer gets rebuilt around that principle.
10. AirMyne — direct air capture
AirMyne builds direct air capture technology using a liquid capture agent and low-temperature thermal regeneration, aiming to lower the energy cost that has historically made DAC too expensive to scale. It's backed by Y Combinator and secured a strategic investment from ENEOS Holdings; a California sequestration site is planned for 2026.
Like Fervo, this is honestly a materials/process engineering company first — AI/ML assists with optimizing the capture-and-regeneration cycle, but the core innovation is chemistry, not a model.
Why it matters: direct air capture is one of the few carbon-removal pathways that doesn't depend on land availability, and lowering its energy cost is the main blocker to scale.
Where AI is the actual product vs. a supporting tool
Being honest about this distinction matters — it's the same discipline our climate-and-AI evidence piece applies to every claim in this space:
| Category | Startups | AI's role |
|---|---|---|
| AI is the core product | Watershed, Kayrros, GHGSat, Overstory, ClimateAi, PassiveLogic, Emerald AI | Pattern recognition, forecasting, or real-time control is the thing being sold |
| AI is a supporting tool inside deep-tech engineering | Fervo Energy, AirMyne | Subsurface modeling or process optimization assists a physical engineering breakthrough |
| AI-native, now part of a larger company | Pachama (Carbon Direct) | Original AI/satellite MRV tech persists inside an acquirer |
What people are asking
Is this list just funding hype, or are these companies actually deployed? Every company here has paying customers, a completed acquisition, an IPO, or a named deployment site in 2026 — not just a seed round and a pitch deck. Fervo's IPO and Watershed's Fortune-500 customer list are the clearest signals of real market validation.
Why isn't [insert well-known climate startup] on this list? This list favors companies where AI/ML is a checkable, real part of the product, following the same two-gate standard the rest of explainx.ai's coverage uses — real traction plus a genuine build/buy consequence for the reader, not just name recognition.
Does using AI make these companies more climate-effective than non-AI climate startups? Not automatically. The measurement bar is the same either way: does it change a real decision or physical system, and does the benefit outweigh the computing footprint. Several companies on this list (Watershed, Kayrros, Overstory) clear that bar because their AI directly produces a verifiable, actioned outcome — a repaired leak, a corrected forecast, a maintained line.
Related reading
- Can AI Solve Global Warming? What the Evidence Says — the evidence-based framework this list applies to real companies
- AI and Climate Change: The Paradox of a Tool That Causes and Fights the Crisis — AI's own energy footprint, set against its climate applications
- Data centers: the real environmental impact (water, electricity) — the infrastructure problem Emerald AI is built to solve
- The AI data-center backlash, mapped — community and grid pushback against AI infrastructure growth
- Algae, trees, and carbon capture: urban air purification — a companion look at carbon-capture approaches outside this list
- Top 10 AI Newsletters to Follow in 2026 — for tracking AI and climate-tech news as it breaks
Official sources: CTVC — H1 2026 climate tech funding · Watershed · Kayrros · Emerald AI · Fervo Energy
Funding figures, valuations, and company status reflect public reporting as of August 2026 and change quickly in this sector — verify current figures on each company's own site before citing them elsewhere.