As far as the White House, Silicon Valley, and Wall Street are concerned, artificial intelligence is a major priority.
In order to support the technology, tech companies are investing billions of dollars in new data centers and infrastructure. The significance of AI to the administration was highlighted in July when the White House released an AI action plan to strengthen American leadership in the field. Nvidia (NVDA) and other AI-related equities continue to set new marks on Wall Street.
However, concerns have been expressed over whether the administration’s policies might undermine its significant AI effort in light of President Donald Trump’s trade war. The price of the components and materials required for supporting certain AI models may increase as a result of certain tariffs.
On August 6, for instance, the president said that he would impose a 100% duty on imports of semiconductors, but he clarified that businesses who had promised to increase their production in the US would be excluded. (He did not specify the precise commencement date of the taxes.) Additionally, he levied a 50% tax on copper in late July. Copper is utilized in printed circuit boards, semiconductors, and other electronic components.
Experts think tariffs won’t hinder technological breakthroughs, despite the fact that they could increase cost uncertainty. This is mainly because the stakes are too great to lose the global AI race.
For big tech firms like Microsoft and Meta, losing in AI would be more costly than any new tariff-related expenses. These organizations probably see the AI surge as a “existential moment” for their operations, according to Dallas Dolen, PricewaterhouseCoopers’ US technology, media, and telecommunications lead.
“If you have the money, cost is not the most important factor that you consider when you’re told it’s an existential threat,” he told.
AI is receiving billions of dollars from Big Tech
A clear message emerged from the late July earnings reports from Google, Microsoft, and Meta: Big Tech is investing heavily in AI, and the results are beginning to show.
For the quarter ending in June, Meta spent $17 billion on capital expenditures, and its profits per share increased 38% over the previous year. Generally speaking, capital expenditures refer to funds spent on infrastructure and data centers, which is perhaps an indication that Meta is increasing its investment in the servers required to support its rapidly expanding AI services.
When Meta (META) released its earnings on July 30, after-hours trading saw a 9% increase, and the company’s shares have now risen almost 30% for the year. Wall Street applauded the findings.
Microsoft’s (MSFT) cloud computing division also contributed to its impressive performance. In its most recent quarter, the firm spent $24.2 billion on capital expenditures, and it aims to invest an additional $30 billion in the upcoming months, the company announced in late July. Following Nvidia, Microsoft became the second business to attain a $4 trillion value last month. The company’s shares have increased by almost 26% this year.
Additionally, Alphabet, the parent company of Google, raised its 2025 capital expenditures to $85 billion due to the increasing demand for its cloud products. In reference to privately owned businesses in the generative artificial intelligence area valued at $1 billion or more, the company claimed that “nearly all gen AI unicorns” employ its cloud services. To date, Alphabet’s stock (GOOG) has increased by around 7%.
Goldman Sachs predicts that the worldwide power consumption from data centers would increase by 50% by 2027 and 165% by 2030 compared to 2023 due to AI, suggesting that this extra infrastructure may be necessary.
Dan Ives, an analyst at Wedbush Securities, stated in a research note after the companies’ earnings results that “we have barely scratched the surface of this 4th Industrial Revolution now playing out around the world led by the Big Tech stalwarts like Nvidia, Microsoft, Palantir, Meta, Alphabet, and Amazon.”
The potential effects of tariffs on the AI push
Estimating the precise impact of Trump’s frequently shifting tariff policy on the cost of constructing and running data centers has proven challenging.
However, according to estimates he’s heard, tariffs might raise building prices by 5% to 7%, said Dolen of PwC. According to the National Association of Manufacturers’ outlook poll, the two biggest business difficulties facing manufacturers in the first quarter of 2025 were rising raw material costs and trade concerns.
However, Michelle Brophy, head of research for tech, media, and telecom at market intelligence firm AlphaSense, stated that “demand is so strong” that large tech businesses are likely to absorb any additional expenses associated with AI infrastructure.
Smaller businesses without billions to spend every quarter have a different situation. Additionally, private investors usually want a quick return on their investment, whereas data centers are long-term investments that may not yield significant returns for years.
According to commercial real estate services firm CBRE, the average time to build a data center was between one and three years between 2015 and 2020. Additionally, Pankaj Sachdeva, a senior partner at McKinsey & Company, stated in October 2024 that a data center is typically usable for 25 to 30 years.
“The degree of uncertainty will have a larger impact in terms of, you know, committing to something that will take multiple years to execute because data centers are long-term projects,” said Laurence Ales, a Carnegie Mellon University economics professor.
Additionally, it’s uncertain if Trump’s tariffs on semiconductors would result in higher data center costs in the future. According to the president, businesses who have “committed” to constructing in the US would not be subject to a semiconductor levy.
At an event on August 6 to announce Apple’s $100 billion plan to manufacture iPhone parts in the US, he stated, “The good news for companies like Apple is, if you’re building in the United States, or have committed to build, without question, committed to build in the United States, there will be no charge.”
Although chipmaking behemoths Nvidia and TSMC have both stated they will increase their US operations, Trump did not specify which businesses would be excluded.
Analysts predict that the White House and Silicon Valley will work together more in the future, which might reduce any possible expenses that tech companies may incur as a result of tariffs. When Trump let Nvidia and AMD to sell their AI processors to China earlier this week, he demonstrated his readiness to engage in negotiations with industry executives by requiring them to give the US government a 15% share in return for export permits. According to reports, the White House is also considering investing in the chipmaker Intel.
The White House’s AI action plan contains legislative proposals for simplifying permits for facilities such as data centers and semiconductor manufacturing facilities, and building AI infrastructure is a major component of this strategy.
According to statistics provided by Statista from Cloudscene, a company that links companies with cloud services, the United States already has more data centers than any other nation. Many of the biggest cloud providers in the world, like Amazon and Microsoft, are American companies.
Matt Pearl, head of the Center for International and Strategic Studies’ strategic technologies program, told that “we need to be mindful that this is an area in which we have an advantage.” “And we’d rather not give that up.”
