Global technology companies invested more than $400 billion in artificial intelligence-related data center infrastructure in 2024, with investment continuing at an accelerating pace into 2026 as the competitive dynamics of the AI industry push major players to outspend rivals in a race for computational capacity and talent. The scale of this investment is unprecedented in the history of the technology industry and is reshaping electricity demand, real estate markets, water consumption, and supply chains for specialized hardware components across multiple continents.

Scale and Scope of Data Center Investment

The United States alone now hosts more than 5,000 data centers, with new facilities under construction or planned in dozens of states. The buildings, which house the thousands of specialized AI chips and associated cooling and power infrastructure needed to train and run large AI models, are enormous structures requiring hundreds of megawatts of electricity each and millions of gallons of water for cooling. Their construction has been transforming local economies in states and counties where large-scale development is occurring, bringing construction jobs and tax revenues while also creating significant demands on local water and electricity infrastructure.

Energy Implications

The electricity demands of AI data centers have become a significant factor in power markets across the United States, Europe, and Asia. Major technology companies are competing for access to both utility-scale electricity supply and dedicated generation capacity, driving investment in new power plants and grid infrastructure. The scale of demand has reinvigorated interest in nuclear power as a reliable, low-carbon baseload source that can serve data center needs without contributing to grid instability or carbon emissions. More than 60 nuclear reactors were under construction worldwide at the end of 2025, partly in response to this demand.

AI Capability Race

The investment surge reflects the extraordinary competitive pressure that AI developers face to maintain or gain capability advantages over rivals. The training of each new generation of large language models requires exponentially more computing power than the previous generation, driving continuous investment in hardware at the frontier of what is technically possible. Companies that fall behind in compute capacity risk being unable to develop the most capable models, which are increasingly seen as strategic assets with implications not just for commercial competition but for national security and geopolitical influence.

Questions About Sustainability

Despite the scale of investment, serious questions remain about its long-term sustainability. The returns being generated by AI applications have not yet clearly justified the extraordinary capital being deployed, creating uncertainty about whether the investment cycle will continue or eventually face a correction. The World Economic Forum's 2026 Global Risks Report identified the potential for an AI investment bubble as one of the significant near-term risks to global financial stability, noting that much of the investment is debt-financed and therefore vulnerable to rising interest rates or disappointing revenue outcomes.