Google is lining up a major bet on AI infrastructure in Texas, backing a multibillion-dollar data-center project leased to Anthropic and operated by Nexus Data Centers. The venture signals a deepening race to host ever-larger AI workloads, with the initial phase reportedly pushing past $5 billion and construction financing being lined up from a consortium of banks. Google is anticipated to provide construction loans, according to Financial Times reporting, with bank financing expected to be arranged by mid-year.

Anthropic recently signed a lease for a sprawling 2,800-acre campus that forms part of its broader infrastructure tie-up with Google. Construction already underway, supported by early-stage debt from Eagle Point, a publicly traded closed-end investment company. When completed, the site is expected to deliver around 500 megawatts of capacity by late 2026—roughly the energy needed to power 500,000 homes—with potential expansion to as much as 7.7 gigawatts. The location sits near substantial gas pipelines operated by Enterprise Products Partners, Energy Transfer and Atmos Energy, enabling the project to leverage on-site gas turbines to support power-intensive compute operations.

Key takeaways

• Google’s financing role and Anthropic’s Texas campus signals a concerted push by a major tech giant to anchor AI compute capacity in the United States, with ft.com reporting a possible >$5 billion initial phase and a mid-year financing close.
• The campus is designed for rapid scale, with a target of 500 MW by 2026 and an ultimate expansion pathway toward 7.7 GW of capacity.
• Energy strategy hinges on proximity to natural-gas pipelines and on-site turbines, underscoring how AI workloads drive sophisticated on-site energy planning.
• A separate legal development places Anthropic at the center of a federal dispute over national-security designations, with a preliminary injunction blocking a Pentagon bid to classify Anthropic as a supply-chain risk.
• Beyond civil-military policy, reports indicate Anthropic’s Claude AI played a role in U.S. military operations, highlighting ongoing tensions between AI deployment, governance, and defense use cases.

A Google-backed AI compute hub takes shape in Texas

The Nexus Data Centers project near Anthropic’s operations represents a notable convergence of cloud-scale AI infrastructure with a leading AI developer. The lease and development plan align with Anthropic’s strategy of building robust compute ecosystems to support its Claude models, while Google’s involvement signals a willingness to finance and potentially co-locate large-scale AI workloads with end users and partners. Construction activity is already underway, bolstered by early-stage debt financing from Eagle Point, and the Financial Times reports that Google will provide construction loans, with a consortium of banks competing to arrange broader financing by mid-year.

The site’s size—about 2,800 acres—reflects the ambition to deploy vast amounts of energy-intensive compute capacity. The initial 500 MW target by late 2026 would position the campus among the larger AI-focused data-center builds in North America, and the prospect of expanding to 7.7 GW underscores the long horizon for AI training and inference workloads as models continue to scale.

Financing race and energy strategy

Financing for projects of this scale is a central part of the story. The Financial Times notes the potential $5 billion-plus initial phase, with Google expected to contribute construction loans and a broader banking syndicate aiming to finalize financing by mid-year. The mix of funding reflects a broader trend in AI infrastructure: large industrial players coordinating capital to accelerate capacity growth, often pairing cloud-scale data-center know-how with strategic partnerships with AI developers.

Energy resilience and reliability are baked into the plan. The Texas site benefits from proximity to major gas pipelines, enabling on-site gas turbines to supply a significant portion of the campus’s electricity needs. This approach not only supports the heavy power draw of AI training and inference but also helps address concerns about cost and grid resilience in high-severity load periods. If the expansion to a handful of gigawatts materializes, the project could become a major anchor for regional energy demand, with potential ripple effects on local economies and energy markets.

Governance and legal tangle around Anthropic

In a separate development that could influence how AI vendors engage with government programs, a U.S. federal judge in San Francisco granted a preliminary injunction blocking the Pentagon from labeling Anthropic a national security risk and halting federal use of its Claude AI. The court’s order paused a Trump-era directive that sought to bar government use of Anthropic’s models, describing the government’s action as arbitrary and cautioning against branding a U.S. company as a threat without a firm legal basis. Anthropic filed suit arguing that Defense Secretary Pete Hegseth overstepped his authority by designating the company as a supply-chain risk through executive action rather than statute.

The ruling adds a nuanced layer to the governance of AI tools deployed in sensitive contexts. While it preserves a pathway for continued federal access to Anthropic’s technology, it also underscores the ongoing policy tension around how such tools are classified, regulated, and used within defense and national-security frameworks. The outcome matters for developers seeking to balance public or defense-related deployments with concerns over safety, ethics, and the prohibition of certain use cases, including lethal autonomous weapons or mass surveillance, a stance Anthropic has publicly affirmed.

AI in defense and broader policy implications

Beyond the court ruling, there are broader questions about how AI models are integrated into defense and government operations. Reports cited by Cointelegraph indicate that Anthropic’s Claude was used by U.S. military units in operational planning for a major strike against Iran, illustrating how AI assistance is increasingly woven into decision-making processes even amid policy debates and regulatory scrutiny. This dynamic highlights the tension between rapid AI adoption in critical domains and the need for clear guardrails, governance structures, and oversight to address safety, privacy, and strategic risk concerns.

For investors and developers, the intersection of giant-capacity AI compute hubs, sovereign policy, and the evolving regulatory environment suggests a shifting landscape. The Texas project exemplifies a future where AI model developers, cloud providers, and financiers collaborate to create purpose-built ecosystems that can meet the escalating demands of next-generation AI workloads. But it also points to potential policy and compliance headwinds that could shape timelines, deployment strategies, and financing terms for similar ventures in the near term.

As this ambitious build proceeds, market participants will be watching several key indicators: milestones in construction progress, the pace and terms of financing from the banking consortium, regulatory developments around government use of AI tools, and the practical outcomes of ongoing governance debates surrounding AI in defense contexts. The convergence of energy infrastructure with AI compute capacity and the regulatory landscape will likely influence how quickly, and at what scale, future AI platforms are deployed across the industry.

What remains uncertain is how quickly the broader ecosystem—cloud providers, AI developers, and regulators—will align to support or constrain such megaprojects. Investors and builders should stay tuned to updates on financing closings, energy-supply arrangements, and any new policy guidance that could affect Anthropic’s deployments and similar AI-enabled data centers in the United States.