Google, Meta and Others Pledge Billions of Dollars to Build AI Data Centers

AWS, Microsoft Slow Down Data Center Deployments

Highlights

Google, Meta, Blackstone and CoreWeave are collectively investing hundreds of billions of dollars to build advanced AI data centers across the United States, with a focus on Pennsylvania.

Google signed a $3 billion hydroelectric power deal to support its $25 billion infrastructure expansion, while Meta plans to launch multi-gigawatt clusters called Prometheus and Hyperion.

CoreWeave and Blackstone-backed QTS are constructing new facilities in Pennsylvania, reflecting a broader trend of pairing AI compute growth with regional energy development.

Google, Meta, Blackstone and CoreWeave are planning to collectively spend hundreds of billions of dollars to build data centers for artificial intelligence, as the arms race heats up to build the computing power needed by this advanced technology.

    Get the Full Story

    Complete the form to unlock this article and enjoy unlimited free access to all PYMNTS content — no additional logins required.

    yesSubscribe to our daily newsletter, PYMNTS Today.

    By completing this form, you agree to receive marketing communications from PYMNTS and to the sharing of your information with our sponsor, if applicable, in accordance with our Privacy Policy and Terms and Conditions.

    Google told PYMNTS Tuesday (July 15) that it plans to spend $25 billion on data centers and AI infrastructure in the PJM electric grid region over the next two years. PJM, a regional transmission organization, coordinates the movements of wholesale electricity in 13 states and Washington, D.C. The coverage area is from New Jersey to North Carolina and Illinois to the nation’s capital.

    Google also inked an over $3 billion agreement with Brookfield Asset Management and Brookfield Renewal to modernize two of Brookfield’s hydroelectric facilities in Pennsylvania, which represent 670 megawatts of combined capacity on the regional grid.

    Google launched an AI training program for workers and small businesses as well. It will start with Pennsylvania.

    “Google’s investments announced today will increase energy abundance and empower Americans with the skills needed to thrive in the AI era,” said Alphabet and Google President Ruth Porat in a statement provided to PYMNTS.

    A day earlier, Meta CEO Mark Zuckerberg wrote in a Facebook post that the company would invest “hundreds of billions of dollars into compute to build superintelligence.” Compute refers to computational power, encompassing chips, servers and connectivity in data centers.

    Zuckerberg said Meta is building several multi-gigawatt clusters, with the first one called Prometheus coming online in 2026. The next one is dubbed Hyperion and will scale up to 5 gigawatts over several years.

    “We’re building multiple more titan clusters as well,” Zuckerberg said in the post. “Just one of these covers a significant part of the footprint of Manhattan.”

    BofA analyst Justin Post said in a research report shared with PYMNTS that Zuckerberg’s post means capital expenditures (capex) on AI are “far from over, and future capex growth is likely.” He raised his projected capex for Meta by $6 billion to $229 billion from 2025 to 2027.

    Post said he also sees Zuckerberg’s post as “reaching out to AI talent, further signaling Meta as a place that is focused on, and will provide resources for, AI innovation.”

    Meta’s data centers will serve the company’s new superintelligence team, and Zuckerberg said in the Facebook post that his goal is to build “the most elite and talent-dense team in the industry.” He reportedly has been poaching talent from rivals, offering compensation of up to nine figures.

    “Meta Superintelligence Labs will have industry-leading levels of compute and by far the greatest compute per researcher,” Zuckerberg said in the post.

    See also: Why Is Silicon Valley Spending a Fortune on AI Data Centers?

    Blackstone and CoreWeave Join In

    The Google announcement was made at the Pennsylvania Energy and Innovation Summit held Tuesday at Carnegie Mellon University in Pittsburgh, which was attended by President Donald Trump and Pennsylvania Gov. Josh Shapiro. Blackstone and CoreWeave also unveiled similar news.

    Blackstone said funds that it manages will invest more than $25 billion to build data centers and power structures in Pennsylvania.

    Data center operator QTS, which Blackstone backs, secured several land sites in Northeastern Pennsylvania to develop data centers. Construction is expected to start by the end of 2028.

    To provide electricity for these facilities, Blackstone said it formed a joint venture with electric utility PPL to invest in new natural gas power generation facilities in the state. Blackstone said Pennsylvania offers low-cost energy and is home to 20% of the country’s natural gas production.

    CoreWeave, which provides AI cloud computing services, announced Tuesday that it is committing $6 billion to equip a new data center in Lancaster, Pennsylvania, the heart of Amish country in the state. The facility will initially offer 100 megawatts, which could be expanded to 300 megawatts. CoreWeave will be the tenant of the AI data center.

    Last month, AWS said it would spend at least $20 billion to expand its cloud computing and AI infrastructure in Pennsylvania.

    For all PYMNTS AI coverage, subscribe to the daily AI Newsletter.

    Read more: