Energy availability has become the single biggest constraint on AI data center growth, forcing major infrastructure investors to rethink how power and computing are built together. Swedish private equity firm EQT Infrastructure VII has agreed to acquire Copia Power from Carlyle Group, a move that underscores how critical energy infrastructure has become to the AI boom.

What Problem Are Energy Campuses Actually Solving?

The traditional model of building a data center and connecting it to the existing power grid no longer works at the scale AI companies need. Interconnection queues at utilities have become so backed up that waiting for grid capacity can delay projects by years. Copia Power has developed a different approach: integrated energy campuses that combine power generation, high-voltage transmission, and data center load at the same physical location and electrical interconnection point.

This integration dramatically shortens development timelines. Instead of waiting for a utility to build new transmission lines and add generation capacity, a company can develop solar farms, battery storage, natural gas plants, and data centers as a single coordinated project. The result is faster deployment and guaranteed power availability for hyperscalers like Meta, Microsoft, and others racing to build AI infrastructure.

Copia currently operates or has under construction more than 2.6 gigawatts of energy generation and storage capacity. The company is also developing more than 9 gigawatts of grid-connected data centers supported by energy campuses that include over 25 gigawatts of solar and storage capacity and 7 gigawatts of natural gas generation assets.

Why Is Private Equity Suddenly Focused on Power Infrastructure?

EQT's acquisition reflects a fundamental shift in how infrastructure investors view the AI opportunity. Energy is no longer a commodity utility; it is now a strategic bottleneck that determines which companies can scale their AI operations and which ones cannot. By acquiring Copia, EQT gains a platform that sits at the intersection of energy and digital infrastructure, two sectors that must now move in lockstep.

The deal also signals confidence that this model will work at scale. EQT already owns a portfolio of digital infrastructure assets including EdgeConneX data centers, Zayo fiber networks, Cypress Creek Energy, and Scale, a company specializing in off-grid data center power delivery. The acquisition of Copia allows EQT to connect these assets across its portfolio, creating an end-to-end solution for hyperscalers and utilities.

"The rapid adoption of AI is changing infrastructure demand and making energy an increasingly important part of digital infrastructure," said Alex Darden, partner and head of EQT Infrastructure Americas.

Alex Darden, Partner and Head of EQT Infrastructure Americas, EQT

How Are Energy Campuses Different From Traditional Data Center Power Models?

The traditional approach treats power as a service purchased from utilities. A company builds a data center, connects to the grid, and pays for whatever electricity it needs. This model assumes the grid has spare capacity and that utilities can quickly add generation. Neither assumption holds anymore in regions with heavy AI investment.

Energy campuses invert this relationship. Power generation and data center load are planned together from the start. A developer identifies a site with good solar resources or access to natural gas, builds generation and storage capacity, and then builds the data center right next to it. The power never has to travel far, transmission losses are minimized, and the utility gains both new generation and new load at the same interconnection point, which is often easier to permit and deploy than adding capacity to an already congested grid.

This approach also gives utilities a clearer path to meet their own electrification and renewable energy goals. Rather than waiting for corporate demand to justify new generation, they can partner with developers who are already building the load.

Steps to Understanding the Energy Campus Model

• Integrated Planning: Power generation, transmission infrastructure, and data center load are designed as a single project rather than separate components built sequentially.
• Reduced Interconnection Delays: By adding both generation and load at the same point, projects bypass many of the queue delays that plague traditional grid connections.
• Renewable Energy Integration: Energy campuses typically combine solar, battery storage, and natural gas generation to provide firm, reliable power while supporting decarbonization goals.
• Utility Partnership: Rather than competing with utilities for grid access, developers work with utilities to add both generation and load, creating a win-win scenario.

What Does This Mean for the AI Infrastructure Race?

The acquisition is expected to close by the end of 2026, and EQT has committed to supporting Copia's management team in scaling the platform and advancing key development projects across the United States. This signals that energy-integrated data centers are not a niche experiment but a core infrastructure strategy for the next decade of AI growth.

For hyperscalers, the implication is clear: companies that can secure dedicated energy capacity will be able to scale their AI operations faster than competitors relying on traditional grid connections. For utilities, the opportunity is equally significant: they can meet electrification and renewable energy targets while supporting the economic growth that AI infrastructure brings.

"EQT's deep infrastructure experience and long-term perspective make it the ideal partner as we continue to scale our platform and develop the energy infrastructure needed to support AI and electrification," said Ray Henger, CEO of Copia Power.

Ray Henger, CEO, Copia Power

The broader context matters here too. Energy availability has become one of the main constraints on data center growth, requiring digital infrastructure and power infrastructure to scale together. This is not a temporary bottleneck caused by a surge in AI demand; it reflects a structural mismatch between how quickly companies want to build AI infrastructure and how quickly utilities can add generation and transmission capacity. Energy campuses are a solution to that structural problem, and EQT's investment suggests the model will define how AI infrastructure gets built for years to come.