Big Tech’s AI spending is skyrocketing It’s not necessarily because companies are building more capabilities.
This week, industry executives pointed to a similar fundamental problem. Rising component costs, particularly memory chips, are leading to increased capital expenditures for data centers and other equipment.
Meta CEO Mark Zuckerberg told analysts on Wednesday: “We are increasing our infrastructure capital spending forecast for this year, much of it driven by higher component costs, particularly memory prices.”
Microsoft echoed this trend, with CFO Amy Hood saying that about $25 billion of the $190 billion in capital spending expected in 2026 is related to component price increases.
Amazon didn’t raise its capital spending outlook at all this week. Still, CEO Andy Jassy said memory costs are “rising” and the company is trying to contain those costs.
Memory prices are soaring as AI demand squeezes supply. Research firm TrendForce predicts that DRAM prices could rise by up to 63% and NAND flash prices by 75% in the second quarter of 2026. NAND and DRAM are different semiconductor technologies used to hold data.
This is reshaping how investors should interpret Big Tech’s spending boom. Let’s consider a theoretical example. If you purchase 100 AI components at $1,000 each, your capital investment will be $100,000. If the cost of each component increases by 25%, each will cost $1,250. The capital investment is currently $125,000; Additional capacity is zero.
Discounting this price impact, Big Tech companies may not have stepped up their AI build plans much this week.
Look at this table for BNP Paribas. Microsoft’s capital spending plans exceeded Wall Street expectations by $32 billion. Of this increase, $25 billion Not ambitious construction, but higher prices. Meta revised its capital spending forecast upward by $10 billion. Rising memory prices could account for most or all of that increase.
RBC Capital analysts discovered this phenomenon in early February. At the time, they estimated that rising memory prices could account for about 45% of total capital spending growth. According to this year’s top cloud providers.
The result is a more nuanced picture of the AI arms race. Investment remains strong, but the headline numbers may overstate the pace of expansion.
Instead, Big Tech companies are increasingly paying a premium to secure rare components, highlighting the price of memory as a key bottleneck in the industry’s next phase of growth.
Sign up for BI’s Tech Memo newsletter here. Please contact us by email. abarr@businessinsider.com.