AI Hardware Imports Mask True Economic Impact of U.S. Buildout
Federal Reserve economists find that reliance on foreign chips and servers creates a statistical disconnect in GDP calculations.
The artificial intelligence investment surge is generating less visible economic growth than the scale of spending suggests, according to new research from Federal Reserve economists.
The disconnect stems from how national accounting treats imports. While companies are pouring billions into AI infrastructure, much of that spending flows to foreign suppliers of chips, servers, and networking equipment — creating an unusual gap between investment activity and its reflection in gross domestic product figures.
Why it matters
As AI becomes central to economic strategy, policymakers and investors need accurate measures of its impact. The current statistical framework may systematically undercount AI's contribution to growth, complicating decisions about infrastructure investment, trade policy, and economic forecasting.
The GDP accounting challenge
Federal Reserve Board economists Paul Soto, Mason Thieu, and Jeffrey Allen examined AI's economic footprint in a paper released this month. They found that AI-related spending — encompassing software, data centers, power infrastructure, and computing equipment — contributed approximately 0.73 percentage points to the economy's 2.1% annualized growth rate in the first quarter of 2026.
However, net imports of computers, peripherals, and parts subtracted 0.45 percentage points from that same quarter's growth. Without the import drag, AI's measured contribution would have been substantially larger.
The researchers noted a fundamental measurement problem: AI lacks a dedicated category in government economic data, forcing economists to piece together its impact from various investment and trade line items.
Import dependency by the numbers
AI-related products represented 23% of all U.S. imports in 2025, up from 15% in 2023, according to research from Minneapolis Fed economist Michael Waugh. Computer hardware accounts for roughly half of these AI-related imports, with the remainder consisting of electrical equipment, networking gear, and cooling systems required for data center operations.
The import offset proved even more dramatic in late 2025. During the fourth quarter, companies accelerated AI hardware purchases, but the GDP boost largely evaporated once trade flows were factored in. The Fed economists estimate the net effect of AI investment contributed just 0.14 percentage points to GDP growth in that period, even though gross spending across relevant categories added about 0.75 percentage points before accounting for imports. Net imports of AI equipment reduced growth by 0.61 percentage points.
Trade policy considerations
The Trump administration has largely exempted many AI-related inputs from broad-based tariffs, implicitly recognizing the sector's dependence on foreign supply chains. This policy choice reflects the tension between domestic manufacturing goals and the immediate needs of the AI buildout.
Looking ahead
Economists project 1.8% annualized growth for the April-June period, with results due Thursday. Goldman Sachs anticipates continued strong business investment driven by AI-related equipment spending, alongside rebounding consumer activity.
The Fed researchers observe that AI's net contribution "varies considerably across quarters, as the drag from net exports of computer, peripherals and parts offsets much of the gross investment in quarters where imports rose sharply."
The findings underscore a broader challenge: as economic growth increasingly derives from technologies built on imported hardware and intangible software assets, traditional measures of economic activity become harder to interpret and may require methodological updates.
These details were first reported by Axios.
This is an original analysis by the Omega editorial team. Source reporting: AI Watch.
Want systems like this working for your business?
Book a Call