Anthropic Invests $50B for U.S. AI Infrastructure, Eyes Texas and New York

Anthropic says it will invest $50 billion in American computing infrastructure, starting with data centres in Texas and New York and adding further sites through 2026. The facilities are being built with Fluidstack and are designed “to maximise efficiency for our workloads,” the company said, with the first sites expected online in 2026.

Jobs, timelines and local impact

Anthropic projects about 800 permanent jobs and roughly 2,400 construction roles across the build programme. The company framed the investment as aligned with U.S. policy priorities: it “will help advance the goals in the Trump administration’s AI Action Plan to maintain American AI leadership and strengthen domestic technology infrastructure,” the announcement added.

Why Anthropic is betting on owned capacity

CEO Dario Amodei framed the investment in scientific and long-term terms: “We’re getting closer to AI that can accelerate scientific discovery and help solve complex problems in ways that weren’t possible before. Realising that potential requires infrastructure that can support continued development at the frontier.” Anthropic says the dedicated sites will support research, enterprise growth and scale for Claude, which it says serves more than 300,000 business customers.

Anthropic chose Fluidstack for its ability to deploy “gigawatts of power” quickly. Fluidstack’s model has been used by other AI customers to accelerate capacity delivery; Anthropic emphasised agility and capital efficiency as guiding principles for the rollout.

Competing with OpenAI’s build-out

The announcement comes amid an intense infrastructure race: OpenAI and hyperscalers have secured multibillion-dollar compute commitments (reported at scale in 2025), prompting questions about power availability, supply chains and whether the U.S. can deliver such large-scale deployments. CNBC reporting noted Anthropic’s move as a direct strategic push to compete with OpenAI’s far larger commitments and to assert a domestic compute footprint.

Large infrastructure programmes raise practical questions: siting, grid capacity, permitting, and local supply chains. Analysts and policymakers have also flagged macro risks—from overcapacity to budget strain on energy systems—as firms race to secure onshore compute. Anthropic says it will prioritise cost-effective and capital-efficient approaches, but execution and timelines will be closely watched.

Anthropic’s $50B pledge signals that AI firms are moving from cloud-first experiments to owning bespoke, high-scale compute estates. If delivered as promised, the build could deepen U.S. compute capacity, create local jobs and strengthen domestic AI sovereignty. But it will also test industrial capacity, local permitting processes and the sector’s ability to scale responsibly.