What are AI data center water constraints? AI data center water constraints are the limits that local water supply, drought risk, and cooling demand place on where AI data centers can be built, how quickly they can come online, and what they cost to operate. The issue moved to the center of the conversation in early June 2026, when Alphabet, Google's parent company, announced plans to raise about $80 billion to fund AI compute and, in the same window, pledged to make its data centers water-positive by 2030. For commercial real estate investors, water is fast becoming the second great constraint on data center development after power. For the broader picture, see our guide to the best AI tools and infrastructure trends for commercial real estate.
Key Takeaways
- Alphabet announced about $80 billion in equity offerings on June 1, 2026 to fund AI infrastructure, including a $10 billion private placement from Berkshire Hathaway.
- Google pledged to replenish 120 percent of the water its data center sites use by 2030, the first major cloud provider to set a water-positive target of this kind.
- Water is emerging as the next siting constraint after power, because data centers use large volumes of water for cooling and often sit in water-stressed regions.
- More than 70 percent of Americans oppose a data center being built in their area, per Gallup, making local water and power impact a real entitlement risk.
- CRE investors underwriting data centers must now diligence water rights, reclaimed water access, and cooling design alongside megawatts and power contracts.
Inside Google's $80 Billion Raise and Water Pledge
On June 1, 2026, Alphabet (Nasdaq: GOOGL) disclosed equity offerings expected to total roughly $80 billion to fund its AI compute buildout, one of the largest capital raises in technology history, anchored by a $10 billion private placement from Warren Buffett's Berkshire Hathaway. The scale drew attention partly because Alphabet already held more than $110 billion in cash, a signal of just how capital-hungry AI infrastructure has become. Days later, Google paired the raise with an environmental commitment aimed squarely at the fastest-growing complaint about data centers: their thirst.
Google pledged to replenish more water than its data centers consume by 2030, targeting 120 percent replenishment at its data center sites. The plan rests on five commitments, including using air cooling in areas with at-risk watersheds, sourcing reclaimed wastewater instead of drinking water, helping utilities upgrade water systems, and continuing annual public disclosure of water use. The gap to close is real. In 2024, Google consumed about 7.2 billion gallons of freshwater and replenished roughly 4.5 billion gallons, or about 64 percent, and the company says its restoration work reached more than 7 billion gallons in 2025 across 165 stewardship projects in 97 watersheds. You can read the industry framing in Axios.
Why Water Is Becoming the New Power
For two years, the binding constraint on data center development has been electricity. Power availability, not capital, displaced location as the number one site selection factor, a shift we detailed in our analysis of the AI data center power crisis. Water is now following the same path. Servers and AI chips generate enormous heat, and water cooling can cut a data center's energy use by roughly 10 percent compared with air cooling, which means operators face a direct trade-off: use more water to save power, or use more power to save water. In water-stressed metros, that trade-off becomes a permitting and reputational issue, not just an engineering one.
The political backdrop is hardening. A recent Gallup poll found that more than 70 percent of Americans oppose a data center being built in their area, and water draw is one of the most emotionally charged objections at local hearings. That community resistance has already blocked or delayed tens of billions of dollars in projects, as we covered in the data center backlash, and it compounds the cost pressure from new large load tariffs that utilities are imposing on big power users.
What It Means for CRE Site Selection and Valuation
Data centers are now commercial real estate's hottest asset class, with the sector requiring up to $3 trillion in investment by 2030 and roughly 100 gigawatts of new capacity expected to create on the order of $1.2 trillion in real estate asset value. Investors increasingly underwrite these assets in dollars per kilowatt per month rather than the traditional sticks-and-bricks approach, effectively pricing megawatts. Water is the new variable in that equation. A site with abundant power but constrained or contested water access carries hidden risk: higher cooling costs, slower permitting, and the possibility that a project stalls at the community level.
For CRE investors, this means diligence has to expand. Two otherwise comparable parcels can diverge sharply once you account for watershed stress, access to reclaimed water, and local sentiment. JLL's research team tracks how power and infrastructure are reshaping these decisions in its 2026 global data center outlook, and water is moving up that list. The asset that pencils today can become a stranded entitlement tomorrow if a drought year turns water permits into a political flashpoint.
How CRE Investors Should Respond
- Add water to your site checklist: Evaluate watershed stress, drought history, and water-rights availability alongside power capacity and fiber access for any data center or powered-land play.
- Favor flexible cooling: Sites and designs that can use air cooling or reclaimed water in stressed regions will face fewer permitting and reputational hurdles.
- Underwrite community risk: Budget time and capital for local opposition, and engage early, because entitlement delay is now a primary cost driver.
- Watch the hyperscalers: When a buyer like Google sets a 120 percent water-replenishment target, expect it to become a procurement standard that flows down to the landlords and developers who serve them.
For personalized guidance on evaluating AI infrastructure opportunities, connect with The AI Consulting Network.
Real-World CRE Applications
Picture a developer weighing two powered-land sites, one in a water-rich Midwestern county and one in an arid Southwestern basin. A year ago the decision might have hinged purely on power timelines and land price. Today the arid site's water question can add months of permitting, force a more expensive air-cooled design, and invite the kind of organized opposition that has stalled projects elsewhere. The water-rich site, even at a higher land cost, may deliver a faster and more certain path to revenue. That is the new calculus. CRE investors looking for hands-on support underwriting data center and AI infrastructure deals can reach out to Avi Hacker, J.D. at The AI Consulting Network.
Frequently Asked Questions
Q: Why are AI data centers a water issue?
A: AI servers and chips generate intense heat, and many data centers use water for cooling. Water cooling can reduce energy use by about 10 percent versus air cooling, but it draws large volumes of water, often in water-stressed regions, which creates permitting, cost, and community-opposition risk for developers.
Q: What did Google actually commit to?
A: Google pledged to replenish 120 percent of the water its data center sites consume by 2030, using air cooling in at-risk watersheds, reclaimed wastewater, utility partnerships, and annual public water-use disclosure. It is the first major cloud provider to set a water-positive target of this kind.
Q: How does water constraint affect data center valuation?
A: Data centers are increasingly valued in dollars per kilowatt per month, which prices power capacity. Water constraints add cost and timing risk through more expensive cooling, slower permitting, and potential project delays, so two sites with identical power can carry very different risk profiles based on water access.
Q: Is the $80 billion Alphabet raise related to the water pledge?
A: They were announced in the same early June 2026 window and reflect the same reality: AI infrastructure is straining both capital and natural resources. The $80 billion funds AI compute buildout, while the water pledge addresses the environmental and community backlash that buildout is generating.