AI Power: Europe’s Power Plant Update, Google’s Curb & Meta’s $2 Billion Sale

Around three weeks ago, United States President Donald Trump announced numerous investments aimed at pushing artificial intelligence (AI) technology.

It includes $25 billion from Google for data centers and infrastructure, $25 billion from Blackstone for data centers and natural gas plants, $6 billion from CoreWeave for data center expansion, and an undisclosed amount from Meta that promised to cover hundreds of billions of dollars for AI data centers for superintelligence.

All this went down at the Inaugural Pennsylvania Energy and Innovation Summit at Carnegie Mellon University, held on July 15, 2025, which was hosted by Senator Dave McCormick and attended by officials from the top tech and energy companies, including Google, Meta, Blackstone, Microsoft, and ExxonMobil.

All in all, the announced investments totaled $92 billion—and covered computing power, infrastructure, and generating sufficient energy to power this dream.

Energy consumption has been a main concern surrounding the development and use of AI. It has also been a major driver directing AI-focused companies to partner with energy companies.

Microsoft, Google, Amazon, Meta, and others—including the U.S. government—have turned to nuclear reactors over the past year to power AI data centers. Data centers form the core of AI development and deployment. They’re specialized facilities that handle the large computational demands of AI workloads. Suffice it to say, they’ve become extremely popular.

Recently, Mayor Patrick Collins of Cheyenne, Wyoming, spoke about a new data center that’s about to begin building in the state’s capital city soon.

A joint project between Tallgrass, an energy infrastructure company backed by Blackstone, and Crusoe, an AI infrastructure provider and data center developer, this data center is not ordinary by any measure. It’s expected to begin with 1.8 gigawatts of electricity and is designed to scale up to 10 gigawatts.

In comparison, 1 gigawatt of electricity is enough to power around 1 million homes.

The interesting bit is that the entire state of Wyoming holds a population of about 590,000 people.

“It’s a game changer. It’s huge,” said Mayor Patrick Collins.

Wyoming is no stranger to data centers. Cheyenne, the capital, has played host to Microsoft data centers since 2012, and even Meta’s $800-million data center, which was announced last year, is close to being completed.


Recently, Meta disclosed plans to partner with external parties to raise money to build a data center. This would be carried out by selling off data center assets worth $2 billion, as the company looks to keep up with growing demands for AI computing power required to fuel its wide range of AI applications and tools.

Reportedly, Meta reclassified $2.04 billion worth of land and construction-in-progress as “held-for-sale”.

The idea is to preserve financial resources by co-developing a host of data centers—and with how AI companies are growing in revenue and popularity, it shouldn’t be too tough to find investors.

“We’re exploring ways to work with financial partners to co-develop data centers,” said Susan Li, Meta’s Chief Finance Officer on a post-earnings conference call. While Li said the company plans to fund most of its capital funding internally, she confirmed that it didn’t currently have any finalized transactions to announce. 

Across the pond, companies from several European countries are looking to upgrade and repurpose old coal and gas power plants by converting them into data centers. Engie (France), RWE (Germany), and Enel (Italy) are adopting this technique to score long-term supply deals to power data centers.

Moreover, leaders such as Microsoft and Amazon are looking to repurpose such sites, gaining access to power grids and water for cooling—resources that are vital but pose challenges in acquiring and concerns in utilizing.

Another major concern related to power consumption in the AI sector is that of putting too much stress on power grids.

To this effect, Google said that it has signed two new utility agreements with Indiana Michigan Power (I&M) and Tennessee Valley Authority (TVA) that’ll help make its data centers more flexible to benefit power grids.

The idea is to reduce or shift its power demand during certain hours or times of the year, especially when demand surges. The move also plays a key role in advancing its 24/7 carbon-free energy ambition, which includes procuring clean energy and supporting the grid through demand-side solutions.

“Technologies like AI are poised to spur a new wave of innovation and economic growth—and meeting AI’s energy needs efficiently and reliably also presents a unique opportunity to modernize our entire energy system,” said Google in the blog post announcing the move. “That’s why we’ve been working to bring flexible demand capabilities into our data center fleet.”

Do you think these moves will shape how technology conglomerates, energy companies, and interested businesses will go about future AI-energy plans, projects, and partnerships?

Let us know in the comments below!