60-Gigawatt Clean Energy Push Fueled by AI Demand

John NadaBy John Nada·Jul 6, 2026·3 min read
60-Gigawatt Clean Energy Push Fueled by AI Demand

AI-driven data centers are sparking a 1970s-scale U.S. power buildout, with utilities expanding to meet surging demand.

The U.S. is gearing up for a massive power infrastructure buildout reminiscent of the 1970s, driven largely by AI data centers requiring long-term power solutions. This ambitious endeavor is underpinned by a significant rise in energy demand, particularly from AI and data centers, which Grid Strategies identifies as the primary force driving this growth. As the U.S. expands its energy infrastructure, major utilities are planning extensive generation and transmission expansions to meet these demands.

According to Yahoo Finance, utilities are planning extensive generation and transmission expansions, with three ETFs—Utilities Select Sector SPDR Fund (XLU), Vanguard Utilities Index Fund (VPU), and First Trust Utilities AlphaDEX Fund (FXU)—presenting diverse investment pathways. Each of these funds provides investors with unique opportunities to engage in this transformative phase. For instance, XLU, which tracks the S&P 500 Utilities sector, amasses a staggering $22.55 billion in assets at a bargain 0.08% expense ratio, offering a low-cost way to own mega-cap utilities that are heavily investing in new infrastructure.

The energy evolution is further highlighted by the EIA's projection of a 50% to 90% increase in U.S. generating capacity by 2050, fueled by the energy-hungry demands of AI and data centers. The EIA's Annual Energy Outlook 2026 emphasizes data center server load as a major incremental contributor to this anticipated growth. This increase in capacity is not just speculative; it's already being set in motion with rapid interconnection processes in regions like PJM, which covers the mid-Atlantic and Midwest and hosts the nation's highest concentration of data centers.

Corporate maneuvers reflect this seismic shift. Constellation Energy's $16.4 billion acquisition of Calpine aims squarely at securing 60-gigawatts of clean energy for hyperscalers. This acquisition is part of a broader strategy to cater to the long-term energy needs of tech giants. Meanwhile, Vistra is doubling down on long-term contracts with tech giants like Meta, echoing a 1990s-style surge in electricity demand—fueled by AI and crypto innovations. Vistra's approach is illustrative of how companies are adapting to the new landscape, framing the current electricity demand trend as reminiscent of the tech-driven surges of the past.

The implications of this energy demand surge are profound. With AI's insatiable appetite driving utility expansions, investors are presented with a field ripe for strategic picks. The right investment fit depends on whether the investor wants the names doing the most generation spending, the widest spread of regulated rate bases, or a tilt toward utilities scoring best on factor models. VPU, for example, offers a broad index spanning large, mid, and small utilities, providing a diversified investment across the sector.

FXU, on the other hand, uses a factor-screening approach that has outperformed cap-weighted peers with 21% one-year returns, although it charges a higher expense ratio, making it more suitable as a satellite rather than a core holding. This fund’s strategy aligns with investors seeking growth and value factors rather than sheer market cap weight, representing a dynamic way to capitalize on the utility expansion spurred by AI demands.

As utilities gear up to meet this increased demand, the role of regulatory frameworks becomes crucial. It remains to be seen how these frameworks will adapt to ensure that the power landscape can support the transformative changes underway. The key question remains: What role will regulatory frameworks play as the U.S. power landscape transforms once again? Investors and industry players alike are watching closely, with the evolving energy demands of AI and data centers setting the stage for a new era in power infrastructure.

Scroll to continue