Major technology companies including Amazon.com (AMZN), Alphabet (GOOG), Meta Platforms, Inc. (META), Microsoft (MSFT), xAI, Oracle (ORCL), and OpenAI have committed to covering their rising power costs. However, investment bank Wedbush cautions that this commitment may create a significant bottleneck and rapidly slow down the construction pace of large-scale data centers.
On March 4, executives from these seven tech firms met at the White House and signed a non-binding five-point agreement titled the "Electricity Costpayer Protection Pledge." The agreement is part of their massive artificial intelligence strategies, where they pledge to secure power supply for the energy-intensive data centers they are building. Specifically, the companies committed to constructing, procuring, or bringing online additional power generation capacity, paying for related infrastructure upgrades, and negotiating separate electricity rate structures with utilities and state governments where the data centers are located.
It was stated that these leading U.S. tech companies have promised to share the electricity costs associated with AI development, which is expected to benefit the public, though it may take some time. The White House noted that the President is calling on the nation's top hyperscale cloud providers and AI firms to build, introduce, or procure all energy required for data center construction and operation, and bear the full cost of their energy and infrastructure needs.
According to Wedbush analysts, led by Dan Ives, while this move addresses a major challenge for the administration ahead of a critical election cycle, it will create an important bottleneck for large tech companies aiming to rapidly expand their data center footprint without impacting profitability. The U.S. is at a pivotal moment in the AI revolution, yet faces significant shortages and challenges in the energy supply needed for data center construction, which could slow down the build-out speed.
The analysts added that despite the administration's push to position the AI revolution as a key pillar of economic growth and national security, the cost of living continues to rise. They highlighted that U.S. electricity prices are projected to increase by an average of 7% year-over-year in 2025, with previous estimates suggesting a 6% rise by 2027 as data center demand outpaces power supply growth.
Ives and his team also pointed out that many communities have already rejected proposals for local data center construction. This is because the costs associated with financing these long-term projects and strengthening the grid are being passed on to American households, even as the companies work on building their own power generation capabilities.