Back to Business & Economics

Anthropic buys $3.4 billion in gas turbines as the AI power crunch bites

Anthropic is spending $3.4 billion on natural gas turbines, a purchase that marks a shift for a company that has so far preferred to rent compute rather than own the physical infrastructure behind it. According to a Financial Times report, the units are Caterpillar generators that the equipment supplier OnePWR had originally ordered. The generators are capable of producing up to 1.7 gigawatts of electricity.

Renting was the plan

Historically Anthropic has leaned on rented capacity. In its S-1 filing the company disclosed that it is paying xAI, the SpaceX subsidiary, $1.25 billion per month through May 2029 to use xAI's computing power, and it said it expects to enter into additional similar services contracts. The turbine purchase breaks with that posture.

The most likely reason is that the market for power has tightened. The backlog for natural gas turbines stretches into the early 2030s, and queues to connect to the electrical grid can run just as long. For a company planning a public listing this year, locking in its own generation removes a variable it does not control.

The transaction also took an unusual path. OnePWR ordered the turbines a couple of years ago but could not find financing to complete the purchase. Anthropic had planned to buy them from OnePWR, and instead went directly to the manufacturer, Caterpillar. The company did not say why it did not buy through the supplier, and it is not clear what OnePWR was missing on financing.

A broader gas binge

Anthropic is not the only buyer. Google, Amazon, Microsoft, and Meta have all announced plans to build large natural gas power plants on their data center sites rather than drawing from the public grid. The push has given a second wind to fracking, the drilling method used to pull natural gas from underground rock.

The scale behind that push is large. If the buildout proceeds as planned, US data centers alone could consume more natural gas by 2035 than Germany and Japan combined. That scale also ties the hyperscalers to gas prices, which are volatile. One analyst expects prices to more than triple by the end of the decade in some markets.

The practical read for anyone planning AI infrastructure is that power is becoming a first-order cost, not a background utility. The labs that can secure generation now are buying it. The rest are waiting on turbine lead times and grid queues that run into the next decade, which is exactly the gap a $3.4 billion purchase is meant to close.

Source: techcrunch.com

More news

Back to Business & Economics

Have something in mind?

Start a conversation