HousingWireHousingWire
The disruptions from rapidly developing artificial intelligence (AI) technology are growing as more businesses and consumers incorporate its use into their daily lives and operations.
But the technology — particularly from its largest providers — is largely fueled by sprawling data centers that have sparked resource concerns. This includes their legions of power-hungry processors, the silicon needed to build them, and the massive water consumption to keep their underpinnings and complex computer systems cool during continuous use.
But another emerging concern is related to commercial real estate. With the data-center sector having faced consistent challenges stemming from the post-pandemic shift away from business in a dedicated office setting, the appetite for the construction of large data centers at a rapid pace could prove consequential for commercial real estate.
This is according to reporting from multiple outlets and a review of various metrics by HousingWire.
Recent reporting from The New York Times detailed some of this activity through its description of Blackstone’s rapid acquisition of AI-centered companies and its purchase of property to support them.
Blackstone spent $10 billion in 2021 to acquire Quality Technology Services (QTS), which “leases its facilities to companies like Amazon and Meta and supplies the electricity and water needed to power and cool their computers,” the Times reported.
But it doesn’t stop with QTS.
According to the report, Blackstone — which already maintained a sizable commercial real estate portfolio of office buildings, warehouses and more — has “sunk more money into data centers and related infrastructure than into almost any other sector in the firm’s 40-year history.”
The total was estimated at more than $100 billion for “buying and lending to data centers, as well [as] investing in construction firms, natural gas power plants and the machinery needed to build them.”
Other prominent firms such as KKR, BlackRock and Blue Owl Capital have collectively spent hundreds of billions of dollars on similar pursuits. This has led to some analysts expressing concern about too many data centers being built to match existing demand.
Michael Elias, an analyst at TD Cowan, spoke about some of this concern in the Times’ report. He warned of potential “oversupply” of such properties, particularly since some prominent technology companies — including Microsoft and Foxconn — have pulled back from leases.
But ChatGPT provider OpenAI recently announced a significant investment to build new facilities overseas, and late last month another prominent investor announced a major purchase for a similar project in Arizona.
According to a January 2025 analysis from JLL, the increasing rate of data-center electrical capacity — estimated at 15% per year — is still not enough to meet demand. Centers will need to increasingly shift to liquid cooling to accommodate needs. That makes the centers themselves more complex to construct, especially when their electricity demands are skyrocketing, according to a separate analysis from Avison Young.
Some companies are already seeing challenges emerge.
In a statement made during the HSBC Global Investment Summit this past March, Alibaba Group Holding Ltd. Chairman Joe Tsai said the demand for AI data centers could be outpaced by supply, at least in the short term.
“I start to see the beginning of some kind of bubble,” Tsai said at the event, according to a Yahoo Finance report. “I start to get worried when people are building data centers on spec. There are a number of people coming up, funds coming out, to raise billions or millions of capital.”
The regulatory landscape could also impact the pace of construction and uptake.
In the House of Representatives, its version of the “Big Beautiful Bill” that was passed — a key Trump administration priority — included a provision that would restrict state-level AI regulation for a period of 10 years. It’s unknown if that provision will survive the Senate reconciliation process.
One House lawmaker who said she was unaware of the provision when she voted for the bill shared outrage and concern.
“Full transparency, I did not know about this section […] that strips states of the right to make laws or regulate AI for 10 years,” Rep. Marjorie Taylor Greene (R-Ga.) wrote in a post on X.
“I am adamantly OPPOSED to this and it is a violation of state rights and I would have voted NO if I had known this was in there. We have no idea what AI will be capable of in the next 10 years and giving it free rein and tying states hands is potentially dangerous.”
Some senators have also expressed misgivings. The razor-thin margins for the Republican majorities in both chambers can survive only so many holdouts if Democrats remain united in their opposition.
State lawmakers have also urged Congress not to restrict them from regulating AI in their own states, according to a letter obtained by The Washington Post. This could impact the broader development of AI, including the construction of data centers within specific localities.
The prohibition on state-level AI regulations may not survive the scrutiny of the Senate, according to The Hill.
