The data centers will be developed in collaboration with Fluidstack, an AI cloud platform known for providing advanced GPU clusters to major technology firms such as Meta, Midjourney, and Mistral. According to Anthropic, the initial sites are scheduled to become operational in 2026, with additional locations planned across the country in subsequent phases. The investment is expected to create 800 permanent technical and support jobs, along with more than 2,000 temporary construction positions during the build-out phase.
CEO Dario Amodei said the decision reflects Anthropic’s commitment to building domestic AI capacity capable of driving innovation within the United States. “We’re getting closer to AI that can accelerate scientific discovery and help solve complex problems in ways that weren’t possible before,” Amodei said. “Realizing that potential requires infrastructure that can support continued development at the frontier. These sites will help us build more capable AI systems that can drive those breakthroughs while creating American jobs.”
The initiative marks one of the largest U.S.-based investments in AI infrastructure to date, positioning Anthropic as a central figure in the nation’s efforts to strengthen technological sovereignty and reduce reliance on foreign data processing capacity. The company’s approach focuses on developing flexible, high-power data environments that can support evolving AI workloads, especially those related to enterprise-level clients.
Anthropic’s expansion comes as OpenAI, one of its main competitors, pursues a similar infrastructure drive. OpenAI has announced over $1.4 trillion in long-term commitments involving chipmakers and major cloud providers, including Nvidia, Broadcom, Oracle, Microsoft, Google, and Amazon. The scale of that spending has sparked debate over whether the U.S. power grid and industrial base can sustain the rapid escalation in compute demand — and whether the current wave of AI investment risks overheating the sector.
Anthropic’s operations have grown rapidly, now serving more than 300,000 businesses globally. Enterprise clients account for a significant portion of revenue, and the number of large accounts generating over $100,000 annually has nearly increased sevenfold within the past year. Internal projections indicate the company expects to reach break-even by 2028 — years ahead of OpenAI, which faces substantial projected operating losses in the same period.
To meet its infrastructure goals, Anthropic selected Fluidstack for its ability to deliver gigawatt-scale data capacity on short timelines. The firm specializes in high-performance computing environments optimized for machine learning and large-scale AI training. Meanwhile, Amazon has already opened a 1,200-acre data center campus in Indiana dedicated to Anthropic’s operations, representing an additional $11 billion investment that is already functional while many competitors are still in early development phases. Anthropic has also expanded its compute partnership with Google, securing tens of billions of dollars in additional resources.
The question of how the federal government might support or regulate AI infrastructure has become increasingly contentious. Last week, OpenAI reportedly asked the Trump administration to extend certain CHIPS Act tax incentives to cover AI data centers and related electrical grid components such as transformers. The request followed public scrutiny after comments by OpenAI’s chief financial officer, who suggested potential federal “backstops” for the company’s massive infrastructure contracts.
Although OpenAI has since downplayed that proposal, the episode underscored broader concerns about how America’s AI infrastructure boom will be financed — and who will ultimately bear the cost. For Anthropic, however, the new initiative represents a strategic commitment to building the physical and technological backbone for the next generation of artificial intelligence, keeping both innovation and employment rooted firmly on U.S. soil.
