Google is preparing to support a multibillion-dollar data center development in Texas leased to Anthropic as competition for AI infrastructure intensifies. Operated by Nexus Data Centers, the project’s initial phase could top $5 billion, with Google expected to provide construction loans, the Financial Times reported, citing people familiar with the matter. A consortium of banks is also vying to arrange financing by midyear.
Anthropic has reportedly signed a lease for the 2,800-acre campus, part of a broader infrastructure agreement with Google. Construction is already under way, backed initially by debt financing from Eagle Point, a publicly traded closed-end investment company. The site is expected to deliver about 500 megawatts of capacity by late 2026—roughly the power needed for 500,000 homes—with potential expansion up to 7.7 gigawatts. Its location near major gas pipelines operated by firms such as Enterprise Products Partners, Energy Transfer and Atmos Energy would allow reliance on on-site gas turbines.
Separately, Anthropic faces regulatory pressure in the U.S. A federal judge in San Francisco temporarily blocked the Pentagon from labeling Anthropic a national security risk and from halting government use of its AI tools. Judge Rita Lin granted a preliminary injunction pausing a directive backed by President Donald Trump that sought to cut off federal use of Anthropic’s chatbot, Claude.
The ruling followed a lawsuit by Anthropic, which argued Defense Secretary Pete Hegseth exceeded his authority in designating the company a supply chain risk. In her decision, Judge Lin described the government’s actions as “arbitrary” and warned against branding a U.S. company a threat without a clear legal basis, saying the measures may have been retaliation for Anthropic’s public stance and could violate First Amendment protections.
The dispute stems from a breakdown in negotiations between Anthropic and the Pentagon over military use of its AI. Anthropic resisted allowing its models to be used for lethal autonomous weapons or mass surveillance, prompting a broader standoff with the administration.
Reports also indicate that U.S. military units used Anthropic’s Claude model during a major airstrike on Iran, even after the administration’s ban order. Military commands, including U.S. Central Command (CENTCOM) in the Middle East, reportedly relied on the model for operational support.
