Goldman Sachs Sees AI Spending Hitting $1.4 Trillion by End of 2027, Nearly 4% of U.S. Economy
Goldman Sachs forecast on Wednesday that annualized spending on artificial intelligence-related hardware and intellectual property will climb to $1.4 trillion by the end of 2027, equivalent to nearly 4% of the U.S. economy, as AI remains the dominant driver of American business investment.
“AI will remain the dominant driver of business investment growth in 2027,” the bank said in its forecast, reported by Reuters on Thursday. Goldman expects spending to top $900 billion by the end of 2026 before reaching the higher level a year later.
A buildout unlike any recent investment cycle
The figures describe an annualized spending pace — the rate at which money is flowing into AI hardware and intellectual property at each year’s end — rather than a cumulative total. Even on that basis, the scale is striking: by the end of 2027, AI-related spending would run at a rate equal to almost one dollar in every 25 generated by the entire U.S. economy.
Goldman estimates that AI-related investment will directly add 5.3 percentage points to capital-expenditure growth in 2027, and 0.6 percentage points to U.S. gross domestic product growth, when semiconductor investment is included. Excluding semiconductors, the bank puts the contribution at 4.3 percentage points for capital-expenditure growth and 0.2 percentage points for GDP growth.
Overall, Goldman forecasts U.S. business investment growth of 8% on a quarterly basis in 2027, nearly double the 4.2% consensus forecast cited by Reuters.
Why the GDP impact looks smaller than the spending
The bank flagged two accounting reasons why the huge spending figures do not translate into an equally large lift to measured economic output.
First, semiconductor investment is undercounted in the official business-investment statistics, Goldman said. Second, most information-processing equipment is imported, so its value is netted out when gross domestic product is calculated. The money is being spent, but much of the hardware is manufactured abroad, which means the spending shows up in the investment data more than it does in domestic output.
That distinction matters for how the AI boom is read. The buildout is real and enormous, but a large share of the value is flowing to chipmakers and equipment suppliers outside the United States, even as the data centers themselves are built on American soil.
The boom keeps getting bigger
Goldman has repeatedly raised its estimates for AI infrastructure spending this year. In September, the bank said the five largest U.S. hyperscalers — Amazon, Alphabet, Microsoft, Oracle and Meta — would spend a combined $1.2 trillion on AI infrastructure in 2027, more than 50% above the roughly $800 billion expected this year and above the Wall Street consensus of about $1.1 trillion, according to Bloomberg.
The spending wave has also reshaped capital markets. Anthropic is seeking a valuation of up to $1.02 trillion in its initial public offering, and companies across the AI supply chain have tapped debt markets to fund data-center construction. At the same time, the sector is under strain: Nvidia shares slid in premarket trading on Thursday as the chipmaker’s board deliberated the future of chief executive Jensen Huang, a reminder that the market’s AI enthusiasm now comes with sharp swings. The Dow Jones Industrial Average has also whipsawed in recent sessions as investors weigh AI valuations against interest-rate and trade risks.
Related Articles
Why this matters
For American businesses, Goldman’s forecast signals that the AI buildout is not a one-year surge but a multi-year capital cycle that will shape hiring, construction, power demand and corporate borrowing through 2027. Data centers need land, electricity and specialized workers, and the companies building them are competing for all three.
For investors, the message is more mixed. The spending supports growth in semiconductors, power equipment and construction, but it also raises questions about returns: companies are borrowing heavily to fund projects whose payoff depends on AI revenue catching up with the investment. Goldman itself has warned that the largest hyperscalers would need roughly $300 billion a year in AI revenue to justify their outlays.
For the broader economy, the forecast underlines a paradox of the AI era. The United States is the center of the AI buildout, yet much of the hardware spending leaks abroad through imports — one reason the GDP boost, at 0.6 percentage points, looks modest next to the $1.4 trillion headline.
Latest developments
The forecast landed on a jittery day for markets, with stocks sliding as Nvidia’s boardroom turmoil weighed on technology shares and traders awaited U.S. jobless-claims data. Goldman said AI would stay the dominant force behind business investment in 2027, even as investors debate whether the spending boom is sustainable.
Watch for the bank’s next update, for earnings reports from the major hyperscalers, and for any sign that AI revenue growth is closing the gap with the capital being poured into the buildout.
Frequently Asked Questions
How much will the U.S. spend on AI by the end of 2027?
Goldman Sachs forecasts annualized spending on AI-related hardware and intellectual property will reach $1.4 trillion by the end of 2027, up from more than $900 billion by the end of 2026.
What share of the U.S. economy is that?
The $1.4 trillion annual pace is equivalent to nearly 4% of the U.S. economy, according to Goldman Sachs.
Why doesn’t $1.4 trillion in AI spending add $1.4 trillion to GDP?
Goldman says semiconductor investment is undercounted in business-investment statistics, and most information-processing equipment is imported, so its value is netted out of gross domestic product. Including semiconductors, AI investment adds an estimated 0.6 percentage points to 2027 GDP growth; excluding them, 0.2 points.
How fast will overall U.S. business investment grow in 2027?
Goldman forecasts 8% business-investment growth on a quarterly basis in 2027, nearly double the 4.2% consensus forecast.
Goldman Sachs’ forecast frames the AI buildout as the defining investment story of 2027: a $1.4 trillion annual spending pace that will drive business investment even as its direct contribution to measured economic growth stays comparatively small. The question now is whether the revenue follows the spending.
Sources: Reuters; Sowetan (Reuters); Superpower Daily; PYMNTS / Bloomberg.
