For much of the past two years, Wall Street has treated AI as the most bullish trade on the board, a growth engine that accelerates earnings, assumes hefty valuations, and promises a productivity windfall sometime down the road.
But the Fed has access to the same numbers and appears more likely to treat the AI boost as a new source of demand in a market still struggling to bring inflation back to its 2% target.
Goldman Sachs currently predicts that AI-related capital investment will approach $800 billion in 2026, and estimates that this rapid increase will raise its full-year business investment forecast to 7.8% and increase capital investment growth itself by about 3.3 percentage points.
TrendForce tracks nine of the world’s largest cloud providers and estimates that total spending in 2026 will reach nearly $830 billion, an increase of about 79% from the previous year. A sizable portion of this increase reflects price increases rather than added capacity, with Microsoft allocating about $25 billion of its $190 billion budget to expensive memory and components.
All of this places considerable weight on the inputs that the Fed tends to watch most closely, which could turn this investment boom into a policy headache.
Where will $800 billion in AI spending actually go?
It helps to imagine this expense in physical terms. All that money comes in the form of land, steel, transformers, copper wire, gigawatt-scale power generation capacity, industrial-scale cooling, and the incredibly skilled and rare professionals hired to put it all together.
Goldman describes this as a wave across servers, semiconductors, memory, power infrastructure, data centers, software and research budgets, and the bank’s long-term model tracks annual capital spending in AI increasing from about $765 billion this year to $1.6 trillion by 2031.
Power is a binding force. In a speech in late May, Federal Reserve President Lisa Cook noted that electricity and water prices have each increased by about 5% over the past year, prices for chips, high-tech equipment and software have all increased, and wages in specialty construction industries have increased significantly. Households are feeling some of that pressure on their monthly bills and are starting to face political backlash as some state legislatures move to slow the development of large-scale data centers.
Central bank leaders have been unusually clear and honest about where this is going. Jerome Powell told reporters in March that the construction fever was “putting pressure on all kinds of goods and services that are needed for construction,” and acknowledged that the effect was “probably pushing up inflation.”
In the same May speech, Cook went further, warning that “the increased investment demand from AI could compound further price shocks,” noting that companies have announced more than $1.5 trillion in data center plans, but only a fraction of them are actually built.
In other words, the demand side of AI appears in pricing data far ahead of the productivity rewards the technology ultimately delivers.
What the Bitcoin interest rate cut bet means
The impact will flow directly from Silicon Valley’s balance sheets to cryptocurrencies. Bitcoin spent much of this year relying on hopes that cooling inflation would give the Fed the freedom to cut interest rates, ease financial conditions and reignite the risk appetite that fueled the 2024 rally.
CryptoSlate has documented how closely this asset tracks liquidity cycles, with its sensitivity overtaking Bitcoin’s halving as the primary price driver. With $800 billion in demand, a rate cut is less likely. Because every dollar of AI-related price pressure gives the Fed one more reason to hold back.
The market is already starting to reassess it. Futures and prediction markets currently have a more than 93% chance of holding the June 16-17 meeting, which will be the first meeting chaired by Kevin Warsh since taking over from Powell in May. CryptoSlate has been tracking the development of the reversal, from a time when bond traders were pricing in year-end rate hikes to an inflation record that froze the Fed.
This repricing also spilled over into the spot price, with Bitcoin briefly falling below $62,000 before falling to around $63,600 by June 4, about half of its October 2025 record and a weekly decline of more than 13%. Much of the damage is due to exits, as the Bitcoin ETF recorded 11 consecutive sessions of outflows worth about $3.45 billion, the longest redemption period since the fund was founded in 2024. Most of that money went directly into the AI and semiconductor stocks that were causing the macro problems in the first place.
Within five years, AI could do what its advocates promise: lower costs, automate routine labor, and reduce inflation by substantially increasing output per worker. However, the construction phase initially tends to work in reverse. When you squeeze years of infrastructure demand into a narrow period of time, price shocks come early and windfalls come late because hardware, energy, and talent are bid up long before real efficiencies emerge.
The gap between immediate impact and delayed benefits is troubling the Fed. Warsh has argued that AI will prove to “structurally eliminate inflation” and usher in “the greatest wave of productivity growth in our lifetimes,” a view that underpins his willingness to lower interest rates. But Cook and Governor Michael Barr disagreed, with Barr flatly saying he didn’t think the AI boom was a reason to cut interest rates.
Traders, on the other hand, have been primarily concerned with timing. Bitcoin, like stocks and other markets, tends to react to the first decision in front of it. So the “productivity theory” that will probably pay off in 2030 is of little use for positions held this week, month, or quarter. With inflation above 3%, Mr. Warsh has little room to act on his convictions in June, no matter where he wants to steer.
The same AI boom that is pumping up technology valuations and driving indexes higher could be exactly what alarmed the Fed and delayed the liquidity cycle that crypto traders have spent 18 months waiting for. If policymakers settle for seeing $800 billion in annual spending as another pillar of persistent demand, Bitcoin’s rate cut trade will be built on a much thinner foundation than its holders would like to admit.
