Corporate America’s fervor for AI investments has dismissed skepticism and concerns about sluggish returns, which were stoked by DeepSeek’s low-cost AI models. The Trump administration’s global trade battle, however, poses a threat to halt the expansion in sectors ranging from energy to software.
Future earnings reports from utilities that supply power to massive data centers like Vistra and Constellation Energy, as well as tech giants like Microsoft and Alphabet, will reveal whether or not the tit-for-tat tariffs between the United States and China are making companies reconsider their ambitious infrastructure plans.
In the face of uncertainty, their clients—which include media outlets, merchants, automakers, and airlines—are cutting down on their expenditure. Analysts saw early indications of tech firms reducing data center contracts and warned that this could have an impact on expenditures in AI tools.
Microsoft and Alphabet’s Google have preemptively reiterated their $155 billion capital expenditure commitments for the year, which is almost half of the $320 billion analysts predict Big Tech will spend on AI this year.
Tech companies are under increasing pressure, meanwhile, as supply chains are disrupted by tariffs, particularly in China. A 90-day tariff reprieve earlier this month did not extend to the world’s second-largest economy, which is essential to the manufacturing of AI hardware.
Analysts predict that if an exemption on electronics is removed, the 145% U.S. tariffs on Chinese goods will significantly raise the cost of data centers.
The majority of the data center equipment and electrical infrastructure are produced outside of the United States. Pat Lynch, executive managing director for data center solutions at commercial real estate services company CBRE, stated that there is frequently a shortage of this equipment and a high demand for it worldwide. This will probably be made more difficult by tariffs, particularly if overseas vendors move this equipment to other markets.
The U.S. economy will be significantly impacted if AI spending declines. Analysts at J.P. Morgan predicted in January that data center expenditures may boost the nation’s GDP by 10–20 basis points in 2025–2026.
The “Magnificent Seven” are a group of high-flying stocks that have driven the market in recent years but have lost around $5 trillion in market value since peaking late last year. Shares of these stocks already bear some of the concern.
Nvidia, a behemoth in AI chips, is down roughly 26% this year after a spectacular stock surge over the last two years that temporarily made it the most valuable company in the world. The value of Alphabet’s stock has dropped by almost 20%.
THE LONG GAME
There have been indications that businesses are delaying the expansion of their data centers. Last month, analysts at TD Cowen reported that Microsoft had canceled projects that were scheduled to utilize two gigawatts of electricity in the U.S. and Europe over the previous six months because of an excess of electricity.
Earlier this month, a senior Microsoft executive stated that while the business may slow down its ambitions, it would keep expanding capacity wherever demand existed.
Planning takes a lot of time and money, and any major new project of size and scope needs to be flexible and improved as we learn and develop with our clients. This means that we are slowing or pausing some early-stage projects, The CEO wrote on LinkedIn.
Amazon.com indicated that the delay in committing to additional data center leases was due to normal capacity management and had nothing to do with its expansion ambitions, according to Wells Fargo analysts on Monday.
In their report, the analysts noted that it does seem that the hyperscalers, or giant cloud companies, are becoming more selective when it comes to renting out massive power clusters.
Data gathered by LSEG indicates that Microsoft is expected to report its slowest revenue growth in seven quarters. Visible Alpha analysis indicated that Azure, the Microsoft company most positioned to gain from AI, is expected to have growth that is more than a year below its previous peak.
Apple, Amazon, and Alphabet are also anticipated to post lesser growth, despite some support from a declining currency.
Nevertheless, some experts and investors contend that even while short-term returns are slow, the long-term potential of AI justifies further investment—a claim frequently made by tech leaders.
In order to stay competitive in the competition to control the technology, Amazon CEO Andy Jassy defended his company’s expenditures in AI earlier this month.
The CEO of private equity company Patriarch Organization, Eric Schiffer, who has boosted his exposure to the “Magnificent Seven” in recent months, stated that the market has greatly underestimated the short-term investment on AI and that this is incorrect. The big IT firms can’t afford to fall behind in the AI race. He predicted that within a year or 18 months, hyperscalers will begin to yield more significant profits.
