Crusoe's $31 Billion Mark Meets a Data Center Backlash
Crusoe raised close to $4 billion at a valuation near $31 billion, and its chief executive says the data center industry's real problem with local communities is communication.
Crusoe CEO Chase Lochmiller said the AI infrastructure company raised nearly $4 billion in fresh funding at a valuation approaching $31 billion, told Bloomberg Tech the company could eventually go public, and argued the data center industry has a "marketing issue" in explaining its water, jobs and tax impacts to local communities.
Crusoe has raised close to $4 billion in new funding, lifting the private AI infrastructure company's valuation to almost $31 billion, chief executive Chase Lochmiller said in an interview with Ed Ludlow on Bloomberg Technology. The money is earmarked for a build-out that spans three layers at once: physical data centers, cloud computing capacity sold on top of them, and managed AI services sold on top of that.
That is an unusually wide remit for a company that is still private. Most firms in this business pick a lane — land and power, or GPUs for rent, or software. Crusoe is attempting to own the stack from the substation to the model endpoint, and the raise is sized accordingly.
A private valuation built on power, not software margins
The thing that separates this generation of AI infrastructure companies from the cloud giants is what the capital actually buys. A software company raising several billion dollars is buying engineers and time. An AI infrastructure company raising it is buying transformers, land, cooling systems, grid interconnection rights and chips — assets that depreciate, sit in specific counties, and draw the attention of the people who live there.
That is the backdrop to Lochmiller's second point. Asked about the mounting local resistance to new data center projects, he did not argue that the concerns are illegitimate. He argued that the industry has a "marketing issue" — that it has failed to communicate what a large facility actually does to a community's water supply, its employment base and its property tax receipts.
It is a notable framing, and it cuts both ways. If the problem is genuinely one of communication, then better disclosure fixes it. If the objections are substantive — that a hyperscale campus consumes water and grid capacity in quantities a county did not plan for, and employs a fraction of the people a factory of comparable capital cost would — then no amount of messaging closes that gap.
Why the opposition is now a financing variable
Local permitting used to be a scheduling detail in these projects. It is becoming a cost-of-capital input. Investors funding multi-billion-dollar campuses need confidence that a site can be energized on a known timeline; a moratorium, a rejected rezoning or a utility rate fight can strand capital that has already been committed to chips and construction.
The tax revenue argument Lochmiller raises is the one that historically wins local votes. Large facilities do generate substantial property and sales tax bases relative to the acreage they occupy, and in rural counties that can be transformative. The jobs argument is weaker on headcount and stronger on construction phases and wage levels. The water argument is the hardest, because it is measurable, visible and local in a way that abstract economic benefit is not.
The industry's response so far has been technical — closed-loop cooling, air cooling in cold climates, reclaimed water agreements. Lochmiller's point is that those responses are not being explained before a project becomes contentious, only after.
The IPO question and what it would expose
Lochmiller also said Crusoe could eventually enter the public markets. He did not commit to a timeline, and there is no filing. But the remark is worth holding onto, because a listing would force into public view the numbers this sector currently discloses selectively: contracted capacity versus built capacity, customer concentration, the depreciation schedule applied to GPUs, and how much of the revenue is genuinely recurring cloud income versus one-off construction.
Private markets have been willing to underwrite AI infrastructure on forward capacity commitments. Public markets apply a harsher lens, and several listed names in the neighborhood of this business — compute providers, power developers, optical and server suppliers — have traded with notable volatility as investors repeatedly re-rate how durable the demand is.
The tape that Crusoe would be listing into
The public comparison set was mixed on the day of the interview. As of the last trade at 19:58 GMT on 18 September 2026, the Nasdaq 100 tracker (NASDAQ: QQQ) was at $721.62, up 0.66% from a prior close of $716.92 and trading near the top of a $715.08–$721.73 day range. The S&P 500 tracker (NYSEARCA: SPY) was at $761.86, down 0.10% from $762.60. The Dow tracker (NYSEARCA: DIA) was at $516.05, off 0.44% from $518.35.
That split — technology firm, the broad market flat, industrials lower — is the pattern that has characterized much of the AI trade. Capital keeps rotating toward companies with a credible claim on AI demand, while the rest of the market absorbs the cost side: higher power prices, higher rates, competition for the same grid.
What to watch from here
- Where the capacity lands. Site selection tells you more about a data center company's real strategy than any funding announcement. Power availability and permitting friction, not proximity to customers, increasingly determine the map.
- Whether the three-layer model holds. Owning data centers, cloud and managed services means competing with your own customers in at least one direction. That tension usually resolves one way or the other.
- Any move toward a filing. "Could eventually" is not a plan. A confidential submission, an auditor change or a CFO hire with public-company experience would be the real signal.
- How the community fight evolves. If utilities and legislatures start ring-fencing residential ratepayers from data center power costs, the economics of siting shift materially — and the marketing problem becomes a pricing problem.
For now, nearly $4 billion at a valuation approaching $31 billion says private capital remains comfortable funding AI infrastructure at scale. The harder test is whether the counties being asked to host it stay comfortable too.
Key facts
- Fresh funding: Nearly $4 billion raised by Crusoe
- Valuation: Almost $31 billion post-raise
- Nasdaq 100 (QQQ): $721.62, +0.66%, as of 19:58 GMT, 18 Sep 2026
- S&P 500 (SPY): $761.86, -0.10%, as of 19:58 GMT, 18 Sep 2026
Frequently asked questions
How much did Crusoe raise and at what valuation?
Crusoe raised nearly $4 billion in fresh funding, which lifted its valuation to almost $31 billion. Chief executive Chase Lochmiller disclosed the figures in an interview with Ed Ludlow on Bloomberg Technology on 18 September 2026. The company remains privately held, with no public listing and no filing announced.
What will Crusoe do with the money?
Lochmiller said the capital will fund the company's push to build AI infrastructure across three layers: physical data centers, cloud computing capacity, and managed AI services. That is a wider remit than most peers, which typically specialize in either land and power, GPU rental, or software services rather than all three.
Is Crusoe going public?
Lochmiller said Crusoe could eventually enter the public markets, but he gave no timeline and the company has not made any filing. A listing would require disclosing details the private market currently does not see, including customer concentration, contracted versus built capacity, and how GPUs are depreciated on the balance sheet.
What did Lochmiller mean by a 'marketing issue'?
He was pushing back on growing local opposition to data centers. His argument is that the industry has failed to communicate its actual impact on water use, employment and local tax revenue, rather than that those impacts are unimportant. In his framing, better explanation ahead of projects would reduce community resistance.
Why does local opposition matter to investors?
Permitting and community resistance have shifted from a scheduling detail to a cost-of-capital factor. Multi-billion-dollar campuses need certainty that sites can be energized on a known timeline. A moratorium, rejected rezoning or utility rate dispute can strand capital already committed to construction and chip purchases.
How were markets trading when the interview aired?
As of the last trade at 19:58 GMT on 18 September 2026, the Nasdaq 100 tracker QQQ was at $721.62, up 0.66%. The S&P 500 tracker SPY was at $761.86, down 0.10%, and the Dow tracker DIA was at $516.05, down 0.44% — technology firm, the broader market flat to lower.
Sources
- Crusoe CEO: Data Center Indusry Has a 'Marketing Issue’ — Bloomberg Technology
