Lead – What: Anthropic is buying $3.4 billion of natural‑gas turbines. Who: The AI startup preparing for a 2026 IPO. When: The deal was reported in November 2026. Where: Turbines will be sourced from Caterpillar in the United States. Why: To secure reliable electricity for its rapidly expanding compute workloads amid a tightening power market. How: By purchasing generators capable of producing up to 1.7 gigawatts, bypassing its traditional reliance on rented compute capacity.
According to TechCrunch, Anthropic – a fast‑growing artificial‑intelligence company that has historically shunned ownership of physical infrastructure – has entered the “natural‑gas frenzy” that is sweeping the AI sector. The firm plans to spend $3.4 billion on Caterpillar‑manufactured generators, a purchase that would give it a combined generation capacity of up to 1.7 gigawatts of electricity.
The acquisition marks a stark departure from Anthropic’s usual operating model. As reported by TechCrunch, the company disclosed in its S‑1 filing in May that it is paying SpaceX subsidiary xAI $1.25 billion per month for computing power through May 2029. The filing also noted that Anthropic expects to “enter into additional similar services contracts,” signalling a continued preference for renting rather than owning compute resources.
Nevertheless, TechCrunch explains that the surge in demand for AI compute has exposed a parallel shortage of reliable electricity. The report highlights that the backlog for natural‑gas turbines now stretches into the early 2030s, while queues to connect new generators to the electrical grid can be equally lengthy. Faced with this uncertainty, Anthropic appears to have opted for a “last‑resort” strategy: buying power generation assets outright.
The turbines were originally ordered by OnePWR, an equipment supplier that had secured the units from Caterpillar a few years earlier. According to TechCrunch, OnePWR was unable to obtain financing to complete the purchase, prompting Anthropic to step in. Rather than acquiring the turbines through OnePWR, Anthropic reportedly dealt directly with Caterpillar, the original manufacturer. The report does not clarify why OnePWR could not secure financing despite Anthropic’s expressed interest, but it suggests that Anthropic wanted to retain the generating capacity for itself rather than risk a competitor snapping it up.
Anthropic’s move mirrors a broader industry trend. TechCrunch notes that other hyperscale players—Google, Amazon, Microsoft, and Meta—have all announced plans to construct large‑scale natural‑gas power plants on or near their data‑center campuses. By generating electricity on‑site, these firms aim to reduce dependence on the public grid, which is often strained by the massive power draws of AI workloads. The report adds that this “binge” has given a second wind to fracking, the drilling technique used to extract natural gas from underground rock formations.
If the current build‑out proceeds as projected, TechCrunch warns that U.S. data centers could consume more natural gas than Germany and Japan combined by 2035. The reliance on natural gas also introduces exposure to price volatility. One analyst cited by TechCrunch predicts that natural‑gas prices could more than triple by the end of the decade in certain markets, a scenario that could leave companies like Anthropic facing “buyer’s remorse” if costs surge after the turbines are installed.
Opinion (Analysis)
The decision to purchase, rather than rent, power generation capacity underscores how the AI arms race is reshaping traditional utility economics. While owning turbines may insulate Anthropic from grid‑capacity bottlenecks, it also locks the company into a fossil‑fuel‑heavy energy mix at a time when many tech firms are publicly pledging carbon‑neutral goals. If natural‑gas prices indeed triple as forecast, Anthropic could find itself paying a premium for both compute and power, potentially eroding the financial advantage it sought through vertical integration.Conclusion – Anthropic’s $3.4 billion turbine purchase signals a pivotal shift in how AI companies address the escalating power demands of large‑scale models. By securing its own natural‑gas generation capacity, the startup aims to mitigate grid‑supply risks, even as industry analysts caution that volatile fuel prices could offset the benefits of such vertical integration. The move also highlights the growing interdependence between the AI sector and the fossil‑fuel industry, a dynamic that may shape both energy policy and corporate sustainability strategies in the years ahead.
Source: TechCrunch