AI and the rise of software investment

Measuring productivity growth is key to understanding how the economy is growing and becoming more efficient. Total factor productivity, which measures the labor and capital used in producing goods and services, increased 1.1 percent from 2019 to 2024. Businesses invest in capital, including computers, machines, buildings, inventories, land, and intellectual property (software, research and development, and artistic originals, like music and art). These investments find their way to new technologies and improved production methods. As businesses adopt these assets, they can produce more output with the same amount of labor and capital, contributing to increases in total factor productivity. Productivity increases come with a multitude of benefits, which may include higher wages for workers, reinvestment in business, increased profits, lower prices, and a better standard of living.

Breaking down investment growth by asset category, 2019–24

The BLS productivity program aggregates individual asset categories of equipment, structures, residential structures, and intellectual property products. These groupings allow us to see how investment shifts over time. Chart 1 illustrates that private business capital investment in the current 2019–24 business cycle outpaced the previous 2007–19 business cycle. However, the increase in investment was not uniform across asset categories. There has been a slowdown in investment in equipment assets, whereas investment in intellectual property products has substantially increased. These trends may be tied to the increased usage of AI between the 2007–19 and current 2019–24 business cycles.

Furthermore, intellectual property products can be dissected into the asset categories of research and development, software, and artistic originals. (See chart 2.) Software investment growth has outpaced investment growth in research and development and in artistic originals between the 2007–19 and current 2019–24 business cycles.

A closer look at the software asset category

The software asset category can be further disassembled into the components of prepackaged, custom, and own-account software.¹ Millions of Americans have used ChatGPT since its deployment in 2022, but AI also includes large language models, generative AI, and much more.² So, how has the rise of AI translated into the business use of software assets? AI, at its core, is software that has been developed to use during the production process and is often provided to customers via prepackaged and custom software. Using industry productivity data, we can see this in publishing industries (except Internet). This industry subsector includes newspaper and book publishers but also includes software publishers who bring the creation of a copyrighted product to market. In 2022, the industry more than doubled its investment in prepackaged and own-account software, and continued sustained investment in the subsequent years of the current business cycle.

Conclusion

As artificial intelligence rapidly transforms technology, businesses are investing more in assets that they hope will bring new efficiency and productivity gains. The rapid growth in software investment from 2021 to 2024 may be one indication of the recent rise in AI usage to assist in production.

To see more details and data related to total factor productivity and AI, check out the productivity page, the productivity and AI page, and the productivity tables page. Total factor productivity data are produced annually by the productivity program.