Anthropic is set to significantly enhance its data center capabilities, enlisting experienced ex-Google executives to spearhead the ambitious initiative. The artificial intelligence company is reportedly aiming to secure at least 10 gigawatts of data center capacity over the coming years, potentially costing hundreds of billions of dollars, according to a report by The Information. As a startup with limited creditworthiness, Anthropic is seeking powerful allies to support this venture, with Google already acting as a guarantor for a data center in Louisiana. The company also anticipates gaining access to up to one million tensor processing units (TPUs) by 2026.
Previously reliant on renting computing power from cloud service providers, Anthropic is pivoting towards leasing its own data centers. To facilitate this shift, the company has hired Tim Hughes from Stack Infrastructure and Brett Rogers, who has extensive experience in building data centers at Google. Winnie Leung, who boasts over 20 years of experience at Google, is also part of the team driving this initiative.
The race for data center capacity is becoming increasingly competitive among AI firms. OpenAI, for instance, announced partnerships last year with major players such as Nvidia and Oracle, aiming for at least 10 gigawatts and 4.5 gigawatts of custom AI accelerators, respectively. The success of these expansive plans—and whether both Anthropic and OpenAI can generate sufficient revenue to support their ambitious infrastructure investments—remains uncertain. However, Anthropic has reported notable revenue growth recently, positioning itself favorably in this competitive landscape.
Infrastructure has emerged as a critical factor in the AI sector, with OpenAI asserting that its data center capabilities provide a significant competitive edge. CFO Sarah Friar underscored the importance of compute power, indicating that the company could have achieved even greater growth with access to more resources, thereby establishing a direct correlation between data capacity and revenue generation.
As the demand for data center infrastructure surges, local communities are increasingly voicing concerns. Microsoft, Anthropic, and OpenAI have pledged to cover the electricity costs associated with their data centers to mitigate the impact on residents facing rising utility bills. Microsoft took the lead in January 2026, subsequently joined by OpenAI and Anthropic in February 2026.
These commitments come amid growing political scrutiny, as regions hosting data centers experience electricity price hikes of 12-16 percent faster than the national average. Lawmakers in states like New York and Washington are drafting legislation aimed at holding AI companies financially accountable for the costs imposed on local communities. The stakes are high: projections indicate that data centers could consume as much as 12 percent of all U.S. electricity by 2028, up from 4.4 percent in 2024.
The unfolding developments in the AI data center landscape underscore the industry’s rapid evolution and the intricate balance between technological advancement and community impact. With significant investments on the horizon, Anthropic and its peers will need to navigate these challenges while striving to maintain their competitive edge in a burgeoning market.
See also
Synolon Systems Unveils AI-Driven Infrastructure to Transform Global Real Estate Transactions
IonQ Acquires SkyWater Technology, Boosting Quantum Computing Manufacturing Capacity
Applied Materials Surges 20% Post-Earnings as AI Chip Demand Drives Growth
Global X Cloud Computing ETF Poised for Growth Amid AI-Driven IT Overhaul
Nvidia Achieves 10x Cost Savings in AI Inference with Open-Source Models on Blackwell
