Anthropic announced Wednesday a massive $50 billion investment in artificial intelligence (AI) infrastructure in the U.S., with plans to construct custom data centers in Texas and New York starting in 2026, with additional facilities expected to follow.
The massive buildout, the latest in a series of frenetic construction projects disclosed in recent weeks by some of tech’s bold-faced names, is in partnership with Fluidstack, an AI cloud platform that offers large-scale GPU clusters to major clients such as Meta Platforms Inc., Midjourney, and Mistral.
Anthropic’s endeavor is expected to generate more than 2,000 construction jobs and 800 full-time jobs at a time when U.S. policymakers are increasingly prioritizing domestic compute capacity and technological sovereignty.
“We’re getting closer to AI that can accelerate scientific discovery and help solve complex problems in ways that weren’t possible before,” Anthropic CEO Dario Amodei said in a statement. “Realizing that potential requires infrastructure that can support continued development at the frontier.”
The announcement comes as Anthropic’s chief rival, OpenAI, pursues its own aggressive expansion. The ChatGPT maker has secured more than $1.4 trillion in long-term infrastructure commitments through partnerships with NVIDIA Corp., Broadcom Inc., Oracle Corp., and major cloud providers including Microsoft Corp., Google, and Amazon.com Inc.
However, the astronomical pace of spending has set off alarms about whether the U.S. possesses adequate power capacity and if the AI sector is in danger of approaching bubble territory.
Anthropic currently serves more than 300,000 businesses, with enterprise clients generating most of its revenue. Internal projections obtained by The Wall Street Journal indicate Anthropic expects to reach profitability by 2028, significantly ahead of OpenAI, which has forecast $74 billion in operating losses the same year.
To support its growth trajectory, Anthropic picked Fluidstack to construct facilities specifically optimized for its AI workloads, citing the firm’s rapid deployment capabilities and ability to deliver gigawatts of power on compressed timelines.
Meanwhile, Amazon has already opened a dedicated $11 billion data center campus for Anthropic on 1,200 acres in Indiana, providing operational capacity while many competitors remain in the planning stages.
Earlier this week, Meta earmarked $600 billion on AI infrastructure projects. The breakneck buildout reflect demand for AI services. “What you’re hearing CEOs say in earnings calls is that they need more data centers and AI chips. If they did, they would have more revenue,” Shiv Ramji, president of Auth0 on AI security at Okta Inc., said in an interview. “AI productivity tools are in demand, and we are seeing breakthroughs in autonomy, robotics, and health care research.”
The infrastructure push arrives amid heated debate over federal involvement in AI financing.
Last week, OpenAI requested the Trump administration expand a key CHIPS Act tax credit to cover AI data centers and grid infrastructure, according to documents obtained by Bloomberg.
The request followed controversial comments from OpenAI Chief Financial Officer Sarah Friar, who suggested during a tech conference of a potential government “backstop” for the company’s compute deals — comments the company later walked back.
The episode has highlighted mounting questions about how America’s AI infrastructure will be funded and whether taxpayers should shoulder any of the financial risk.
