Four broad investment categories associated with AI accounted for an estimated 39% of U.S. real GDP growth in the first three quarters of 2025, according to a 2026 analysis by the Federal Reserve Bank of St. Louis. That is a substantial contribution—but it is not an official measure of AI’s share of growth. The categories include spending that is not exclusively AI-related, and their contribution fell sharply by the third quarter.
What the 39% estimate measures
The St. Louis Fed estimated that four investment categories contributed 0.97 percentage points to real GDP growth during the first three quarters of 2025—equivalent to 39% of measured growth over that period. The categories were information-processing equipment, software, research and development, and data-center construction.
As an Amazon Associate I earn from qualifying purchases.
This is an accounting estimate for a group of AI-associated investments, not a tally of spending on AI alone. Official statistics do not classify a distinct AI industry, and the national accounts do not have a line item that isolates AI’s contribution. The proxies include non-AI activity; for example, R&D spending is not all AI-related. The St. Louis Fed’s data-center calculation also uses an imputed value for September 2025.
What’s actually slowing this PC down?
Pick the symptom - the matching free tool is one click away.
That distinction matters: the estimate says how much these measured investment categories added to GDP growth during a period. By itself, it does not establish that AI caused every dollar of spending, that the spending will pay off, or that AI has already increased economy-wide productivity.
#1 Best Overall
The contribution changed over the course of 2025
The same four categories added less to GDP growth in each successive quarter, according to the St. Louis Fed’s 2026 estimates. Their contribution can slow even while investment levels remain high: GDP growth accounting reflects changes in spending, not simply how large the existing investment stock is.
| Period | Contribution to real GDP growth | Share of GDP growth |
|---|---|---|
| 2025 Q1 | 1.30 percentage points | Not stated by the St. Louis Fed in the figures reported here |
| 2025 Q2 | 1.16 percentage points | 30% |
| 2025 Q3 | 0.48 percentage points | 11% |
| 2025 Q1–Q3 combined | 0.97 percentage points | 39% |
The quarterly figures and the combined-period figure describe different aggregations; the combined estimate should not be read as a constant quarterly share. The falling quarterly contribution also makes “AI-related investment drove a large share of growth” a period-specific statement, not a claim that the share remained high throughout the year.
Big corporate spending is related, but it is a different measure
The Federal Reserve Board reported aggregate 2025 capital expenditure of $412 billion for Amazon, Google, Meta, Microsoft, and Oracle, equal to about 1.31% of U.S. GDP. That figure illustrates the scale of investment by five large companies, but it is not the St. Louis Fed’s GDP-growth contribution estimate. The corporate capex series excludes leases, while the Fed’s proxy-category calculation covers broader national investment categories.
Nor does a capex total translate directly into an equal addition to GDP growth. The measures have different coverage and answer different questions: one totals reported capital spending by selected companies; the other estimates the contribution of selected investment categories to the change in real GDP.
Investment growth is not the same as productivity growth
The Council of Economic Advisers reported that U.S. investment in information-processing equipment and software grew at a 28% annual rate in the first half of 2025, compared with 5.5% in 2024. This measure is broader than AI, and its growth rate describes investment—not the productivity gains businesses may or may not achieve by using the equipment and software.
BEA industry-level production accounts offer another related view: growth in information-industry capital was among the top contributors to real GDP growth from 2021 through 2024. BEA says that pattern is consistent with significant contributions from investment in AI and AI-related services, but also notes that AI production and services span multiple sectors. The information industry is therefore only one of the primary industries associated with AI.
Independent reader supportYour contribution helps us test, update, and keep practical guides available for everyone.What adoption and productivity evidence can—and cannot—show
Business adoption surveys
Federal Reserve Board data from the revised Census Business Trends and Outlook Survey put the share of businesses using AI at about 18% at the end of 2025, with about 21% planning to adopt it. Census changed the survey question in November 2025: it began asking about AI in any business function rather than AI used to produce goods or services. Readings across that change are not perfectly comparable, and adoption rates alone do not show whether AI raised output or productivity.
Free tools Windows power users keep installed
One-click scans. No signup required.
BEA’s October 2026 research spotlight draws on a different measure: Gallup survey data on workers. It reports that the share of workers saying they used AI at all rose from roughly 20% in mid-2023 to nearly 50% by early 2026; frequent use rose from roughly 10% to more than 25%. Those worker figures should not be treated as the same series as the Census business-adoption readings: they survey different populations and measure use differently.
Best Value
Early evidence on output and employment
In the October 2026 spotlight, BEA found stronger post-2020 real-output trajectories in state-industry groups where workers reported more AI use. Employment differences were generally positive, though estimated less precisely. This pattern is more consistent with output growth alongside stable or stronger employment than with a simple account in which AI adoption is already producing broad employment displacement.
The finding is an association, not a clean causal estimate. BEA cautions that reverse causality, small samples, and the fact that adoption is not random make it difficult to identify AI’s independent effect. More productive or fast-growing workplaces may be more likely to adopt AI, as well as potentially benefiting from it.
A separate BEA working paper’s baseline model finds evidence consistent with AI enhancing productivity and saving inputs. But an alternative specification, using a different timing assumption for when AI became pervasive, produces less robust results. It is early evidence with meaningful sensitivity to model choices, not settled proof of economy-wide gains.
Do these 3 things before closing this tab:
1Repair Windows errors before they cause bigger problems2Scan for outdated or missing drivers - takes under a minute3Clear out junk files and repair common Windows errorsHow to read the headline claim
- What is well supported: broad investment categories plausibly linked to AI made a large estimated contribution to U.S. real GDP growth in the first three quarters of 2025.
- What the 39% does not mean: it is not an official AI-only share, a measure of AI’s contribution to productivity, or proof that AI caused all the spending counted.
- What remains open: how much of the buildout will translate into durable productivity and output gains, and how much of those gains can be causally attributed to AI.
The most accurate shorthand is that AI-related investment categories accounted for a large but declining share of measured growth in early 2025. That describes the investment buildout without mistaking it for proof that AI has already transformed the U.S. economy’s productivity.
Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
