Artificial
intelligence has become one of the dominant investment themes of this decade.
Technology companies are spending billions of dollars on data centers, advanced
semiconductors, software development, and AI infrastructure. The scale of these
expenditures has led many investors to ask an important question:
How much of the recent strength in U.S. investment can be
attributed to the AI buildout?
Answering
that question is more challenging than it may initially appear. The national
income accounts, which contain detailed data that make up Gross Domestic
Product (GDP), do not contain a separate category labeled "AI
investment." As a result, there is no single economic data series that
allows us to directly measure the amount of capital being allocated toward
artificial intelligence.
Nevertheless,
the national accounts provide several clues.
Total Business Investment Has Accelerated
I
began by examining Real Private Nonresidential Fixed Investment, a broad
measure of business capital expenditures that includes equipment, structures,
and intellectual property.
Using
2010 as a base year of 100, business investment rose from:
- 100 to 144 during the 2010-2014 period
- 144 to 173 during the 2015-2019 period
- 170 to 239 from 2020 through 2026
These
figures indicate
that the current investment cycle has
been considerably stronger than the years immediately preceding the pandemic:
While
this observation is noteworthy, it does not by itself demonstrate that AI is
responsible. Business investment encompasses a broad range of activities,
including manufacturing facilities, industrial equipment, energy
infrastructure, transportation networks, and traditional technology spending. To
determine whether AI-related activities are playing a larger role, it is
necessary to look beneath the surface.
Investment Is Increasingly Concentrated in Intellectual
Property
The
first category I examined was Intellectual Property Products, which includes
software, research and development, and other intangible assets.
When
indexed to 2010, the data show the following:
- Intellectual property investment rose from 100 to 132 during 2010-2014
- It increased from 132 to 190 during 2015-2019
- It then surged from 193 to 310 between 2020 and 2026
The
acceleration since 2020 is difficult to ignore:
More
importantly, intellectual property has been growing faster than overall
business investment. When I calculated intellectual property investment as a
percentage of total nonresidential fixed investment, the share rose steadily
over time. As noted in the graph below, by 3Q2026, intellectual property accounted
for 43.7% of total nonresidential fixed investment, up from roughly the low-30%
range a decade earlier.
This suggests that businesses are allocating a growing share
of their investment budgets toward software development, research, and other
intangible assets.
The Hardware Side Tells the Same Story
AI
is not solely a software story. Modern AI systems require massive computing
resources, advanced semiconductors, networking equipment, servers, and data center
infrastructure. To capture this dimension, I examined Information Processing
Equipment and Software investment. The results were even more striking.
Using
2010 as a base year, the data show the following:
- The index increased from 100 to 144 during 2010-2014
- It rose from 145 to 212 during 2015-2019
- It exploded from 212 to 432 between 2020 and 2026
In
real terms, this category more than doubled over the post-pandemic period:
Furthermore,
when measured as a share of total nonresidential fixed investment, information
processing equipment and software rose from
roughly one-quarter of business investment in the early 2010s to 45% by 2026:
This represents a profound shift in the composition of
capital spending: An increasing portion of America's investment budget is being
directed toward computing equipment, software, and digital infrastructure rather
than traditional forms of capital investment.
The Evidence Points to a Structural Shift
No
individual data series can prove exactly how much of this growth is attributable
to artificial intelligence. After all, these categories also include cloud
computing, cybersecurity, enterprise software, digital transformation
initiatives, and other forms of technology investment. However, the timing is
difficult to dismiss.
The
most pronounced acceleration in intellectual property investment and
information processing equipment investment occurred during the same period
that AI emerged as the dominant technology investment theme. It coincided with
unprecedented spending on data centers, AI infrastructure, advanced
semiconductors, and large-scale model development by companies such as
Microsoft, Amazon, Alphabet, and Meta. More
importantly, the two categories most closely associated with AI are not merely
growing rapidly. They are consuming a steadily larger share of the nation's
overall investment budget.
Conclusion
The
national accounts do not provide a direct measure of AI investment. Yet they
reveal something equally important: where businesses are choosing to allocate
their capital.
Since
2020, the composition of U.S. business investment has shifted decisively toward
intellectual property products, information processing equipment, and software.
These are precisely the categories one would expect to benefit most from the
buildout of artificial intelligence.
Viewed
through this lens, the AI boom is not simply a stock market phenomenon or a
story confined to a handful of technology companies. It is increasingly visible
in the nation's investment data itself. Firms are allocating a growing share of
their investment budgets toward both the "brains" of the AI
ecosystem, such as software and research, and the "hardware" that
powers it, including computing equipment and digital infrastructure. While the
precise magnitude of AI investment remains difficult to measure, the direction
of travel is becoming increasingly clear: AI is claiming a larger share of
America's capital spending, and with it, a larger role in shaping the future
path of economic growth.







