Power gets all the headlines in the artificial intelligence infrastructure debate.
Every analyst tracks gigawatt availability and utility connection timelines.
But a silent, arguably more severe, existential threat is looming over hyperscale data center expansion: Water.
A standard 100 MW AI data center consumes over a million gallons of water a day for its evaporative cooling towers.
As next-generation hardware densities push past 100kW+ per rack, the sheer volume of water required to reject that thermal load into the atmosphere is hitting a hard wall.
Municipalities and utility providers in major compute hubs are actively denying permits for new builds due to local water scarcity.
AirJoule Technologies ($AIRJ offers a paradigm shift:
They pull pure distilled water directly out of thin air, using the data center’s own low-grade waste heat to power the process.
If this technology scales successfully, an AI data center becomes practically water-independent. Hyperscalers could build massive compute clusters in arid environments without touching local municipal water supplies, completely bypassing the regulatory permitting bottleneck.
1. The Technical Moat: Low-Temperature Sorbent Regeneration
This is not a giant consumer dehumidifier.
Traditional atmospheric water generation requires massive amounts of electricity to cool air down to its dew point, making it economically unviable at an industrial hyperscale level.
AirJoule’s competitive advantage comes from its proprietary Metal-Organic Framework (MOF) technology.
These advanced materials act like microscopic sponges, capturing large amounts of moisture thanks to their extremely high surface area.
At the heart of the technology is an efficient process that uses small temperature differences to repeatedly absorb and release water, enabling low-cost and energy-efficient water production.
I won't go into the technical details, as they are not essential to the investment thesis. What matters is that the process is highly complex and difficult to replicate, giving AirJoule a unique technological moat.
2. The Product Roadmap: What They Sell and When
AirJoule does not make any revenue yet, but they are transitioning from testing their technology to actively selling it. They are splitting their business into two main product lines:
1. AirJoule Prime
- What it is: A massive industrial system built to generate about 2,000 liters of pure water every single day.
- Where it stands: The first full-scale unit has already been built at their Delaware facility.
- The Catalyst: They are scheduled to deploy this pilot system at the Net Zero Innovation Hub for Data Centers in Fredericia, Denmark, in Summer 2026. They will be collaborating directly with Google, Microsoft, Data4, and Schneider Electric. This is the ultimate test. If it proves it can cool a real data center efficiently, it gives big tech companies the "green light" to buy it at scale.
2. AirJoule Core
- What it is: Smaller, plug-and-play units built for local use that generate about 250 liters of water per day. It comes in two versions:
- The Water Generator (Core AWG): Launches in late 2026. It is designed for military bases (using waste heat from trucks and generators to make drinking water for soldiers) and homes in drought-heavy areas.
- The Dehumidifier (Core DH): Launches in 2027. Instead of making drinking water, this version simply strips humidity from industrial warehouses. Early tests show it uses 40 percent less energy than standard industrial air conditioners, slashing electricity bills.
3. The Power Players: Tier-1 Institutional Validation
For a micro-cap company, AirJoule commands an unusually dense web of global industrial joint ventures and partnerships.
This level of backing heavily mitigates the "science project" risk:
- GE Vernova: A formalized 50/50 Joint Venture (AirJoule, LLC) gives them exclusive manufacturing and supply rights for climate control and atmospheric water harvesting across the Americas, Africa, and Australia.
- Carrier Global: A strategic development partnership to integrate AirJoule’s core dehumidification IP directly into commercial HVAC lines, taking aim at a massive global air-conditioning footprint.
- BASF: A scaling partnership focused on the chemical production of the engineered super-porous MOF materials at true industrial volumes.
- CATL: A 50/50 Joint Venture (CAMT Climate Solutions) to manufacture, deploy, and commercialize the technology within the rapid-growth Asian markets.
- TenX Investment: An exclusive regional distribution agreement covering six Gulf Cooperation Council (GCC) countries, targeting sovereign AI infrastructure builds in the ultra-arid Middle East.
4. Balance Sheet Health & Runway Stress-Test
Following a $22M net equity raise in January 2026 and a subsequent $15M registered direct offering in late May 2026, AirJoule has aggressively shored up its liquidity.
Because they currently generate zero commercial revenue, valuing the stock via traditional Forward P/S or P/E metrics is impossible. Instead, the valuation rests entirely on cash preservation versus time-to-market.
Financial Runway Scenarios (Based on $45M Combined Liquidity)
With roughly $45 million in combined liquidity following their recent $15 million direct offering, valuing the company rests entirely on cash preservation versus time-to-market. We can evaluate their runway across three distinct scenarios.
The Base Case: Fully Funded into 2028
Aligned with current management guidance, the company expects an annual cash burn of $25 million.
Drawing down on the $45 million liquidity pool provides a runway of roughly 21 to 24 months. This fully funds operations into early 2028. This scenario carries a low dilution risk and successfully bridges the company through the pre-revenue phase, securing the commercial launch of both their Core and Prime platforms without forcing an early capital raise.
The Accelerated Burn Case: Mid-2027 Capital Requirement
If supply chain chokepoints or material scale-up costs hit their manufacturing pipeline, annual cash burn could jump to $35 million.
Under this heavier spend rate, the runway shrinks to approximately 15 months, extending only to mid-2027. This carries a moderate dilution risk, forcing management back to the equity markets before their recurring revenue ramp fully materializes.
The Severe Commercial Lag Case: High Dilution Risk
In this scenario, annual burn remains perfectly controlled at $25 million, but strict data center design-in lifecycles push initial recurring revenue out to 2029. While the current liquidity ensures the hardware reaches commercial status, the delayed revenue timeline would force a heavily dilutive equity raise in late 2027. This is the highest risk scenario for current shareholders—the technology works perfectly, but the timeline mismatch forces painful dilution before profitability is reached.
5. The Management Team: Who They Are And Skin In The Game
AirJoule’s leadership looks more like an industrial commercialization team than a typical early-stage startup, with strong ties to major strategic partners and meaningful insider alignment.
Matt Jore – CEO & FounderFounder of AirJoule and the main architect of the business. He brings 30+ years of experience in product commercialization. Importantly, he owns about 11% of the company, which is a large insider stake and means his financial outcome is closely tied to the company’s success.
Pat Eilers – Executive ChairmanFounder of Transition Equity Partners and a long-time energy transition investor. He also led the SPAC process that brought AirJoule public. While his skin in the game(6.7% of the company) is mainly through his firm and leadership role rather than a disclosed large personal stake, he is deeply financially and strategically involved in scaling the company.
Bryan Barton – Chief Commercialization OfficerPreviously held a senior role at GE Vernova, focused on ventures and incubation. He now leads commercialization at AirJoule, including working closely with GE on their joint venture. His alignment comes more from execution role and career transition than direct ownership disclosure.
Strategic Links (Carrier connection)The board includes Ajay Agrawal, who is also Chief Strategy Officer at Carrier Global, his presence strongly aligns AirJoule with Carrier’s long-term strategic interests in HVAC and cooling infrastructure.
Bottom line
The key point is alignment:
- The founder has large direct ownership (~11%)
- Key executives come from GE and Carrier ecosystems
- The board includes strategic industry insiders, suggesting strong industry buy-in rather than a purely academic startup structure
6. The Bear Case: Super High-Execution Challenges
While the tailwinds are undeniable, an investor must account for structural hardware risks:
- The 24 to 36-Month Data Center Design-In Lag: Hyperscale data centers are planned, permitted, and architected years in advance. Even if Microsoft, AWS, or Google run pilots of AirJoule Prime and love the data, the hardware integration lifecycle for entirely new facilities takes years. Material revenue contribution from the data center vertical is unlikely to materialize before 2028.
- Hardware Margin Cap: Management is targeting a mature gross margin profile of roughly 35%. While proprietary, this remains a physical, capital-intensive manufacturing business. It will never command the high-80% gross margins or the 20x software multiples of pure-play AI IP licensing firms.
7. The Verdict: Is it a Commodity?
Absolutely not.
The core innovation is the ability to regenerate a highly engineered MOF using low-grade waste heat from data centers, a capability protected by a substantial IP portfolio and years of technical development.
This is not commodity technology.
There is a reason industry leaders such as GE Vernova and Carrier chose to enter exclusive partnerships rather than build competing solutions from scratch. Their involvement speaks to the uniqueness and potential value of the technology.
My Take: The $AIRJ Thesis
AirJoule ($AIRJ) is essentially a leveraged bet on the most overlooked bottleneck in the AI infrastructure buildout: water.
The company has strong founder-led management with skin-in-the-game and is very strategically structured.
Until the company physically proves its technology at scale, the stock will likely trade in a volatile, limited range.
It is impossible to assign it a precise valuation today, but the core catalyst is undeniable. If this technology allows hyperscalers to bypass local municipal water permits, which is currently a massive hurdle for new construction, it drastically accelerates a data center's time-to-market. That alone is a massive green flag.
Look at the fuel cell industry as a parallel.
Before the AI data center boom, fuel cells were a slow-growing niche. Today, Bloom Energy ($BE) has exploded into a $94 billion company, growing revenues over 130% year-over-year strictly by solving the data center power bottleneck.
Water is the exact same type of existential chokepoint.
If AirJoule successfully scales this solution, reaching a valuation greater of $10 billion is entirely realistic.
A Critical Caveat:
Unlike established technologies, this is brand-new hardware.
Fuel cells like those from Bloom Energy ($BE) had been in development for decades before the AI boom hit, AirJoule is still in the early stages of proving its manufacturing and durability at scale.
There is a meaningful "execution risk" here that investors should not overlook.
Ultimately, if the technology works, AirJoule likely won't even stay independent.
Commercial success makes them an immediate acquisition target for a hyperscaler or an industrial HVAC giant looking to monopolize a critical piece of next-generation infrastructure.
TLDR;
What do I like?
- The founder has large ownership (11%)
- The management is strong and strategic
- They aren't doing it alone. They have deep-rooted JVs and partnerships with titans like GE Vernova, Carrier Global, and BASF
- They are solving the "water bottleneck" for data centers, which is arguably the biggest physical constraint on the AI boom right now.
- They have a concrete pilot program in the EU with the world’s largest tech hyperscalers.
What do I not like?
- They are currently a pre-revenue company. It is impossible to know exactly how quickly they can convert "pilot programs" into actual, recurring cash flow
- Unlike software, which can be fixed with a patch, hardware requires physical manufacturing. If they face supply chain chokepoints or material defects, they cannot just "pivot" overnight.
- Because they are a pre-revenue startup, if the tech fails to perform at scale or their primary partners move on to a different solution, the stock has no fundamental floor and could realistically go to zero
- Data centers are planned years in advance. Even if the tech is perfect, it could take 24–36 months before they see meaningful, large-scale revenue hit the balance sheet.
- Unlike fuel cells, which have been around for decades, this is a new hardware. We don't have years of long-term durability data to prove these systems will last a decade in harsh desert or industrial environments
So as of today, I don't really have a reason to invest in these "zero or hero" stocks.
The upside can be enormous, but so is the risk, and at this stage I'd rather focus on companies with clearer execution paths and more predictable outcomes.
Maybe that changes in the future.
If the company matures, starts commercializing successfully, and proves the business model works, I'll happily revisit it, it's still super early.
Disclaimer: As of the time of writing, I do not hold any $AIRJ shares
