US GDP Growth Slows to 1.5%: Is AI Spending Hiding a Stronger Economy?
The 1.5% figure is annualised. It shows what growth would look like if the second quarter's pace continued for a full year. The economy actually expanded by around 0.4% from the first quarter.
The more important detail is that consumer spending grew 3.4%, while business investment excluding housing increased 8.5%. Real final sales to private domestic purchasers, which combines consumer spending and private fixed investment, grew 4.2%.
This measure removes government spending, trade and inventory changes. It therefore gives a clearer view of demand from households and private businesses.
Is AI Spending Hiding a Stronger US Economy?
Partly, yes.
The 1.5% GDP figure is accurate. However, rising imports and lower government spending weakened the headline even though private demand remained strong.
US imports increased at a 12.5% annualised rate and subtracted around 1.64 percentage points from Q2 growth. Some of that increase came from capital goods such as telecommunications equipment, semiconductors and industrial machinery. These products are also used to build AI servers and data centers.
Not every imported chip or machine was purchased for AI. The entire import drag should not be described as an AI effect. Government spending, exports, inventories and non-AI imports also shaped the final GDP number.
The more accurate conclusion is that an import-heavy AI buildout was one reason strong business spending did not fully appear in headline GDP. To understand why, it helps to look at how GDP is calculated.
GDP measures goods and services produced inside a country, not everything American consumers and companies purchase.
The basic calculation is:
GDP = Consumer spending + Investment + Government spending + Exports - Imports
Suppose an American technology company buys an AI server worth $100 from overseas.
Illustrative transactionGDP calculationAI server recorded as business investment+$100Imported equipment adjustment-$100Direct US GDP contribution from the server$0US construction, software and installationPositive contribution
The company spent $100 and increased its computing capacity. But because the server was made outside the US, the foreign-produced value does not count as US GDP. The domestic construction, software, engineering, power and installation connected to it do add to US output. This is why the wider group of companies supplying AI data centres matters to the economic impact of the AI boom.
Think of it as an American restaurant buying an imported oven. The oven may help the restaurant serve more customers later, but the oven itself was not produced in America. The local installation work and the additional meals produced later add to US GDP.
How Strong Is the AI Investment Boom?
The latest company results show that AI demand is still growing much faster than the wider economy.
AI indicatorLatest figureChange or guidanceNvidia quarterly revenue$96.2 billionUp 106% year on yearNvidia Data Center revenue$89.0 billionUp 117% year on yearNvidia next-quarter revenue guidance$108.0 billionCompany forecastAlphabet Cloud revenue$24.8 billionUp 82% year on yearAlphabet Q2 capital expenditure$44.9 billionMainly technical infrastructureMeta Q2 capital expenditure$31.1 billionIncluding finance-lease payments
Sources: Nvidia's fiscal Q2 2027 results, Alphabet's Q2 2026 earnings call and Meta's Q2 2026 results.
Nvidia's revenue rose 106% and its Data Center business grew 117%, while Alphabet and Meta continued spending heavily on infrastructure. A detailed breakdown is available in INDmoney's Nvidia Q2 earnings analysis.
These figures are global, not entirely AI-related and not entirely located in the US. They should not be treated as direct additions to US GDP. They do, however, show that the AI investment cycle remains strong.
The Three Ledgers Behind the AI and GDP Puzzle
The apparent contradiction becomes easier to understand when AI's impact is separated into three ledgers.
LedgerWhat it measuresWhat the latest data showsCorporate ledgerRevenue, profits and capital spendingAI demand remains strongGDP ledgerValue produced inside the USImports weakened reported growthProductivity ledgerAdditional output created using AIBenefits are appearing gradually
The corporate ledger is already showing a boom. The GDP ledger counts only value produced inside the US and removes imported value. The productivity ledger moves slowest because businesses need time to turn AI into higher revenue or lower costs.
A Federal Reserve Bank of Atlanta study, based on a survey of nearly 750 executives, found that more than half of companies had invested in AI.
The surveyed companies expected AI to improve reported labour productivity by around 3% in 2026. The improvement implied by their expected revenue and employment was lower, at approximately 1.9%. These are expectations, not confirmed national productivity figures, but they show the likely delay between buying AI and seeing its full economic benefit.
Does the 1.5% GDP Rate Signal a Recession?
The latest report does not show an economy in recession.
GDP remained positive, consumer spending grew 3.4%, private domestic demand increased 4.2% and real gross domestic income rose 2.2%. The BEA's measure of corporate profits from current production also increased by $400.9 billion during the quarter.
However, the slower headline should not be ignored. If weakness spreads to employment, incomes and consumer spending, the economic outlook could deteriorate.
Inflation is another risk. The Q2 PCE price index increased at a 5.3% annualised rate, while core PCE increased 3.6%. These are quarterly annualised rates, not year-on-year inflation figures. Persistent inflation could keep interest rates higher and make financing more expensive. INDmoney's US inflation analysis explains how inflation, interest rates and technology stocks are connected.
The 1.5% figure is the BEA's second estimate. The third estimate and annual data update are scheduled for September 30, 2026.
What Does the GDP Report Mean for AI Stocks?
The GDP report does not directly weaken Nvidia's current demand story. Nvidia's revenue, Data Center growth and guidance give a more direct view of the AI infrastructure cycle.
For investors tracking AI stocks, the larger risk is indirect. If inflation keeps interest rates high, investors may place a lower value on future technology earnings. If the broader economy weakens, businesses may eventually delay AI projects.
The long-term test is whether companies earn acceptable returns from their AI spending. If spending rises faster than the financial benefit, capital expenditure could eventually slow.
What Would Show That AI Is Strengthening the Wider Economy?
AI must move through three stages before its full impact becomes visible in GDP:
Build: Companies purchase chips, servers, software and data centers.
Adopt: Businesses integrate AI into normal operations.
Produce: AI helps companies generate more output using the same resources.
The US is clearly deep into the build stage. Adoption is spreading but remains uneven. The economy-wide production benefit is only beginning to appear.
The biggest economic breakthrough will not come when Nvidia sells another AI chip. It will come when thousands of businesses use that computing power to produce more goods and services at a lower cost.
Author's Take: The Headline Is Weak, but the Economy Underneath Is Stronger
AI spending is partly hiding a stronger economy, but not because the GDP calculation is wrong.
The headline was pulled down by imports and government spending even as consumer spending and private investment remained strong. America's AI buildout also depends on imported equipment, so some of the spending does not count as domestic production.
AI is still in its expensive foundation stage. Companies are spending heavily on chips, data centers, power and software, while the wider productivity benefit is arriving more slowly. The key question is whether that infrastructure produces lasting productivity growth before the spending cycle slows.
The AI revolution is not missing from the GDP data. It is appearing in a form that the 1.5% headline does not fully capture.
Source: IND Money