Key Highlights
- Anthropic has secured over 12 preliminary data center lease agreements in the United States, representing more than 1 gigawatt of total power capacity
- The AI firm is negotiating with Google to serve as a financial guarantor for these substantial lease commitments
- The company is transitioning from cloud-based infrastructure rental to operating proprietary server facilities
- A recent Series H funding round brought in $65 billion, pushing the company’s valuation to $965 billion
- The firm has submitted confidential IPO documentation and may go public at a $1 trillion market cap
As Anthropic gears up for a potential stock market debut, the artificial intelligence company is making strategic moves to own and operate its computing infrastructure. The startup has secured preliminary lease commitments for more than twelve U.S. data center facilities, collectively offering power capacity that surpasses 1 gigawatt.
Google is currently negotiating to provide financial backing for these lease commitments. Having a $2 trillion tech giant guarantee the agreements would provide data center landlords with substantial assurance regarding payment obligations.
This represents a significant operational pivot for Anthropic. Until this point, the company has followed the typical AI startup playbook by purchasing computing resources from major cloud service platforms like Amazon Web Services and Google Cloud.
By securing and managing its own data center space, Anthropic aims to eliminate intermediary costs. Operating proprietary hardware infrastructure typically offers substantial cost advantages compared to purchasing capacity through cloud service providers.
Google’s Multifaceted Partnership
The connection between Anthropic and Google extends far beyond a conventional business relationship. Google functions simultaneously as an equity investor, cloud infrastructure supplier, and prospective lease guarantor for Anthropic.
This past April, Alphabet announced a commitment of up to $40 billion in Anthropic. Google has additionally collaborated on designing specialized server processors that Anthropic may utilize within these upcoming facilities.
Anthropic has been increasing its deployment of Google’s proprietary TPU processors and is reportedly preparing to scale operations to approximately 1 million chip units.
This infrastructure buildout follows Anthropic’s November 2025 announcement regarding a $50 billion investment in American data center development. The newly signed lease agreements indicate these expansion plans are actively progressing.
Public Market Debut Approaching
Anthropic completed a $65 billion Series H financing round recently, with Altimeter Capital, Dragoneer, Greenoaks, and Sequoia Capital leading the investment. This funding round established a company valuation of $965 billion, positioning Anthropic as the highest-valued AI startup, surpassing OpenAI.
CFO Krishna Rao stated the capital would enable the company to “serve the historic demand we are experiencing, stay at the research frontier, and bring Claude to more of the places where work happens.”
The company has submitted confidential registration documents for a United States initial public offering. Specific details regarding share pricing and offering size have not been disclosed.
Should Anthropic achieve a $1 trillion valuation upon listing, it would join the ranks of the world’s most valuable corporations. The offering could potentially become one of the two or three largest IPOs ever recorded, trailing only SpaceX and Saudi Aramco.
Just days ago, Anthropic introduced Claude Fable 5, representing its most advanced consumer AI model to date. According to the company, Fable 5 demonstrates superior performance in programming tasks, research applications, and sophisticated long-form content generation.
Anthropic isn’t the only player pursuing infrastructure expansion. OpenAI, xAI, and numerous other AI companies are actively competing for data center resources. Developing cutting-edge AI models requires massive computational capabilities, which in turn necessitates power-intensive facilities.
The industry-wide competition to secure infrastructure resources is intensifying as artificial intelligence services experience continued demand growth.
