A New Power Player Is Emerging In Energy: Big Tech (Photo by Sean Gallup/Getty Images)Getty ImagesA fundamental shift is underway in the power sector: the companies that consume the most electricity are starting to dictate how it gets built.Driven by artificial intelligence, data center demand is rising at a pace the grid was never designed to accommodate. In 2024, U.S. data centers consumed roughly 4% of the nation’s electricity. By 2030, that share is expected to double.That scale of demand is forcing a structural change. After years of navigating a complex and slow moving utility system, technology companies — particularly hyperscalers like Microsoft, Amazon and Google — are no longer waiting. They are stepping into the power sector, with capital and urgency shaping how and how quickly new infrastructure gets built.From Buyers To BuildersTechnology companies are moving upstream into the power sector itself. Not long ago, power purchase agreements were considered cutting edge for large electricity buyers. Today, hyperscalers are driving grid buildout with power infrastructure investments that were unheard of just a few years ago. Last December, Alphabet paid $4.75 billion to acquire Intersect Power, a deal that included multiple gigawatts of power. Microsoft has contracted to restart the Three Mile Island reactor in Pennsylvania. NVIDIA’s venture arm, NVentures, joined a $650 million funding round for TerraPower, which is developing an advanced nuclear reactor in Wyoming.These investments signal that the line between hyperscalers and energy companies is beginning to blur. Tech is becoming energy.MORE FOR YOUA New Class Of Power PlayerCOLUMBUS, OH: AI-ready data center located on a seven-acre campus (Photo by Eli Hiller/For The Washington Post via Getty Images)The Washington Post via Getty ImagesIn the age of AI, compute and power are no longer separable. The result is the emergence of a new class of power player not defined by generation assets alone. Rather, hyperscalers are motivated to enter the power industry due to the scale and urgency of their own demand and the need to ensure reliability of the grid around them. The power sector, shaped by utilities and power producers, operates on long, regulated timelines. Infrastructure is built through deliberate planning and governed by predictable returns. Growth is incremental with caution always at the helm. The model is no longer viable under the strain of AI. With speed-to-power a driving force of the industry, gigawatt scale data centers are materializing on timelines far faster than the utility system was designed to accommodate. Google: Supply and DemandThis dynamic is propelling hyperscalers to step into the gap. Earlier this month, Google announced construction of a 1GW data center campus near Detroit, Michigan. Alongside the announcement the company committed to bring 2.7GW of net new power to the regional grid. Adding significantly more power than its own facility will consume, Google is effectively acting as a grid developer to ensure the broader infrastructure can meet its demand and reliability needs.Utilities remain essential operators of the grid but the pace and direction of development are increasingly driven by large, concentrated sources of demand. With the ability to deploy capital quickly, hyperscalers have stepped into a new role as power leaders, influencing how projects are initiated and financed. CERAWeek: The Super Bowl OF EnergyNowhere is this shift toward a new class of power players more visible than at CERAWeek, which kicks off today in Houston. Long known as the "Super Bowl of energy," CERAWeek has been the gathering place of energy incumbents for decades. This year, the roster looks different. Companies like Microsoft, Google, and NVIDIA are taking center stage alongside the oil majors as emerging power infrastructure players. Sessions like "Powering the AI Economy" and "Data Centers and the Grid" reflect a new reality: the largest demand for power is the demand for computeNvidia CEO Jensen Huang shows new AI chips during the Nvidia GTC (Photo by Justin Sullivan/Getty Images)Getty ImagesAI is binding digital and physical infrastructure together. Companies like NVIDIA act as catalysts for how quickly demand materializes. Breakthroughs in compute are immediately reflected in power requirements. The scale and speed of that technology development is actively shaping infrastructure decisions. The effect is already visible in how projects are planned and executed. Data center development is influencing transmission, generation and siting decisions. Federal and state policy is racing to keep pace. The traditional boundaries between the technology sector and the energy industry are giving way to a new paradigm: a tightly coupled system where each shapes the other. How quickly the grid can keep up may largely depend on who builds it.