🔍 Read the full analysis: What Happens When AI Spending Slows? A Look At US Growth on ThorstenMeyerAI.com
Get the latest gadgets delivered free with Prime
- Fast, free delivery on millions of items
- Prime Video, Amazon Music and more included
- Member-only deals all year
TL;DR
The Bureau of Economic Analysis estimated that US real GDP grew at a 3.4% annual rate in the third quarter of 2026, above forecasters’ 3% expectation. Computers and software contributed 1.09 percentage points, but imported computer equipment also weighed on GDP; the figures do not isolate AI or establish what a slowdown would do.
The US economy grew at a 3.4% annual rate in the third quarter of 2026, according to an advance estimate released October 29 by the Bureau of Economic Analysis (BEA), with computer equipment and software adding 1.09 percentage points to growth. The figures show a large role for technology investment, but the government does not report a separate AI category, and imported equipment complicates estimates of how much that spending added to production inside the United States.
The BEA’s estimate put growth above the 3% forecast in a FactSet survey cited by the Associated Press, and up from 2.2% in the second quarter. The 3.4% figure is an annualized pace: the economy’s actual growth during July through September was 0.8%. The report is an advance estimate, based partly on assumptions where September data were not yet available.
Consumer spending contributed 2.52 percentage points, with households spending more on health care, restaurants and hotels, and prescription drugs. Fixed investment contributed 1.66 points and inventories 1.61 points. Government spending added 0.90 points. Exports subtracted 0.68 points, while imports subtracted 2.64 points; imports are deducted in GDP because they represent goods produced abroad.
Within business investment, computer and peripheral equipment spending rose at a 66.7% annual rate and contributed 0.72 percentage points. Software added 0.23 points, while other information-processing equipment added 0.14. Together, those categories contributed 1.09 points, about 32% of GDP growth. That is an estimate based on broad categories, not a government measure of AI spending specifically.
Infographic 01 · US GDP, July to September 2026
Where the 3.4% came from
Consumers carried most of the growth. Equipment, software and a big inventory build added the rest, and a surge in imports took a large slice back.
Contributions to growth, in percentage points
Each bar is one kind of spending. Bars right of the line added to growth, bars left of it subtracted. Together they make the 3.4%.
- Consumer spendinghealth care, restaurants and hotels, medicines+2.52
- Equipmentof which computers and peripherals +0.72+1.16
- Software and researchintellectual property; software alone +0.23+0.47
- Business buildingsstructures+0.15
- Housingresidential investment−0.12
- Inventoriescompanies stocking up+1.61
- Governmentmostly federal+0.90
- Exportsled by petroleum products−0.68
- Importssubtracted in GDP; led by capital goods−2.64
Sum: +3.4% (details do not add exactly because of rounding)
Advance estimate, the first of three; the second follows on 25 November 2026. “Annual rate” means the quarterly pace extended to a full year.
Sources: BEA, GDP advance estimate Q3 2026, 29 Oct, Table 1.5.2 (contributions, expanded detail) · AP, 29 Oct (consensus).
Powered by Thorsten Meyer AI
The Import Cost Behind AI Investment
The figures matter because they show both the scale of technology-related purchases and a limit on what those purchases say about domestic growth. The BEA estimates that business purchases of computers added 0.72 percentage points, but imports of computer equipment also count against GDP when the equipment was made abroad. The source article’s calculation estimates imported computer gear subtracted about 1.1 points from growth, more than the direct contribution from business computer purchases. That is an author’s estimate, not a separately published BEA measure of AI’s net effect.
A slowdown in spending would not translate mechanically into an equally large drop in GDP. If businesses bought fewer imported servers, imports could also fall, reducing the subtraction. But a pullback could hit particular firms and locations more sharply than the national figure suggests, including equipment suppliers, construction contractors, power projects and lenders exposed to data-center development. The actual effect would depend on which purchases stopped, where equipment was produced, and whether other spending replaced them.
The labor picture also differs from the investment figures. The source reports 29,000 jobs added in September, compared with an average of 45,000 a month over the previous year. That contrast does not prove technology investment has failed to create jobs; the available figures do not establish a direct causal link. It does show why GDP growth, construction activity and employment should be read as separate measures.
Popular right now
Vitamix Propel Series 750 Professional-Grade Blender, 4 Programs, Black
De'Longhi 15-Bar Pump Espresso Machine with Milk Frother & Temperature Control
Dyson V8 Cyclone Cordless Vacuum
As an affiliate, we earn on qualifying purchases.
Data Centers and the GDP Measures
GDP measures production within the United States, not the nationality of a company or the intended use of a product. The BEA reports quarterly growth at an annual rate and also calculates how much different categories contributed in percentage points. Its accounts contain categories such as computers, software and structures, but no dedicated AI line. Treating all computer purchases as AI investment would overstate what the data can establish.
Separate construction data offer a view of the physical build-out. The Census Bureau put private data-center construction at an annual rate of $85 billion in August, up 73% from a year earlier. Factory construction fell 20% over the same period. These figures describe construction spending, not the share of GDP growth caused by AI, and they cover a different period from the third-quarter GDP estimate.
Computer equipment and software spending has also risen over a longer period. The source’s calculations from BEA tables put combined spending on information-processing equipment and software at an annual rate of $1.77 trillion, or 5.3% of GDP, in the third quarter. The same categories accounted for an average of about 45% of growth over the four quarters ending in September, according to the author’s arithmetic. That estimate uses broad categories and should not be read as a direct measure of AI’s contribution.
“Forecasters had expected growth of 3%.”
— FactSet survey cited by the Associated Press
What the Advance Estimate Cannot Show
The October 29 release is the first of three GDP estimates and may be revised as more data arrive. The BEA said some figures were based on assumptions because September source data were unavailable. The second estimate is scheduled for November 25.
It is also unclear how much of the computer and software spending was specifically for AI. Those categories include other business technology, and the national accounts do not identify servers used for AI separately. The source’s estimate of the import drag is its own calculation; it is not a BEA estimate of AI investment’s net contribution.
Nor do the cited data establish that AI spending caused the reported employment or stock-market movements. The source says the Nasdaq fell 2.9% on the Friday after the GDP release, but that market change alone does not show why investors sold shares or whether spending by technology firms is slowing. The scale, timing and economic effects of any slowdown remain uncertain.
Revisions and Spending Trends Ahead
The next scheduled GDP milestone is the BEA’s second estimate on November 25, which will incorporate additional information and may revise the third-quarter figures. Later estimates will provide further updates. Readers can compare those revisions with subsequent data on business investment, imports, construction and employment rather than treating the advance estimate as final.
Evidence of a sustained slowdown would require more than a market decline or commentary about investor concerns. Relevant indicators include computer-equipment and software purchases, data-center construction, orders and imports of capital goods, and hiring in related construction and supplier industries. Those measures cover different parts of the economy and may not move together.
For now, the confirmed picture is strong third-quarter growth alongside substantial investment in computer equipment and software, with a large offset from imports. The next estimates may change the size of those contributions, while future data will be needed to show whether technology spending is slowing and how widely any pullback affects businesses, workers and consumers.
Key Questions
How fast did the US economy grow in the third quarter of 2026?
The BEA’s advance estimate put real GDP growth at a 3.4% annual rate for July through September. The actual increase during the quarter was 0.8%.
How much did computers and software contribute?
Computer and peripheral equipment contributed 0.72 percentage points, software 0.23 points, and other information-processing equipment 0.14 points. Together, those categories added 1.09 percentage points, about 32% of reported growth. They are not a direct measure of AI alone.
Did AI investment account for all of that technology spending?
No such breakdown is available in the GDP report. The BEA reports categories such as computers and software but does not separate AI-related purchases from other uses.
Why do imports affect the estimate?
GDP counts production inside the United States, so imported equipment is subtracted from the spending totals. As a result, purchases of imported servers can raise business investment while their import value reduces the measured contribution to US production.
When will the GDP estimate be updated?
The BEA’s second estimate is scheduled for November 25, 2026. It may revise the advance figures as more source data become available.
Source: ThorstenMeyerAI.com
