The complex corporate web behind a $3.2 billion AI data center
Ars Technica Petala Ironcloud ● Covered by 2 sources
A fire hit a half-built AI data center in New York, and crews say they went in almost blind. The $3.2 billion site has a weird chain of owners, tenants, and backers behind it.
Based on reporting by Ars Technica, Petala Ironcloud — read the original for the full story.
Summary, retelling and take written by AI under human oversight; images are AI-generated illustrations. How we work · Report an error
A fire in early June at the Lake Mariner data center in Somerset, New York, did more than damage an unfinished building. It exposed how little the local fire department appears to have known about the place when it rolled up to the scene. Crews reportedly found no working alarm, no suppression system, and three dead hydrants. The safety documents firefighters were supposed to have access to were said to have burned in the blaze too.
Steve Matisz, chief of the Barker Fire Department, said his team went into the building “kind of blind.” They were facing heavy black smoke from chemicals they couldn’t identify because the sheets meant to explain the hazards were apparently gone. “It’s been a difficult situation,” Matisz said, and he said he wasn’t sure what to make of the claim that the documents burned up.
The site itself adds another layer of weirdness. It sits on a former coal mine on Lake Ontario, and the project behind it is not a simple one-company build. TeraWulf owns and operates the data center, but it does so on land it leases from a company owned by its own CEO.
Then there’s Fluidstack, a UK-based AI company that will run the center. Google has warrants for a future 14 percent equity stake and has agreed to guarantee Fluidstack’s lease payments. Anthropic is among the AI companies that the facility is meant to help serve. That is a lot of moving parts for a place where local firefighters say they showed up without the basic information they needed.
My take — AI-written commentary, not fact-checked reporting
This is the kind of corporate nesting doll that keeps showing up around AI infrastructure: everyone is involved, nobody is plainly in charge, and the local fire crew gets the worst of it. If a $3.2 billion project can’t keep its safety paperwork and alarms straight, the glossy AI talk is doing a lot of heavy lifting. The industry loves scale; it should try competence sometime.
Read more about this at: Ars Technica
