brainstorming in the modern office

Catalyze economic growth through capital investment in GenAI

Boost to global GDP growth from increased investment in GenAI R&D, infrastructure, software creation and company adoption will be significant.


In brief

  • To establish GenAI as a cornerstone of modern industry, substantial capital investment will be required.
  • Investing in AI could boost global GDP by 0.5% to 1% by 2033, representing between US$500b and US$1t.
  • Unlocking the productivity potential of GenAI requires deployment of tangible (infrastructure) and intangible (e.g., technology and skills) investments.

In this installment of our series on the economic impact of AI, we delve into the realm of capital investment in generative AI (GenAI). As GenAI has emerged as one of the key components of economic impact, business leaders today find themselves at a crossroads. The October 2023 EY CEO survey indicates a striking dilemma: while a significant 62% of business leaders acknowledge the urgency of acting on GenAI to prevent competitors from gaining a strategic edge, an almost equal percentage (61%) express reservations due to the uncertainties surrounding the formulation and execution of an AI strategy.

The survey further reveals an “adoption paradox.” It highlights that two-thirds of organizations that have successfully launched at least one AI initiative anticipate that AI will revolutionize their entire business and operational models within a mere two-year span. In contrast, organizations with more extensive AI experience, defined as those having completed five or more AI-related initiatives, project a more cautious timeline of three to five years for AI to wield similar transformative effects.

This disparity in expectations underscores the presence of “unknown unknowns” in AI adoption, particularly in determining the nature and extent of capital investment required for laying a robust AI foundation.

In assessing the potential economic impact of GenAI from a capital investment perspective, we examined the near-term boost to growth from increased investment in research and development, infrastructure, software creation and company adoption. Drawing parallels with the IT revolution in the period of 1980-2000, our two main findings are:

  • Significant boost to demand: Assuming trend growth around 8.5% in investment categories where GenAI will be most significantly captured, we estimate that capital investment in GenAI will contribute about 0.1 percentage points (ppt) to US GDP growth annually over the next five years. Our baseline, however, is that business investment will likely be 25% faster, leading to an incremental boost to short-term growth of 0.1 ppt of GDP per year, worth over US$150b after five years. A more optimistic scenario could see 50% faster business investment growth, leading to an incremental boost to short-term growth of 0.2 ppt of GDP per year, worth a cumulative US$300b by 2028.
  • Long-term boost from supply: In our baseline where business investment is 25% faster than the current trend growth, the potential growth rate of the economy would rise by 0.1 ppt per year in the 2028-33 period, lifting real GDP by nearly 1% over the baseline by 2033, or the equivalent of a US$250b boost over a decade. Assuming capital investment in AI technology grows 50% faster than the 2017-22 trend pace over the next five years, the annual capital contribution to long-term GDP growth in the 2028-33 period would rise by 0.2 ppt. This stronger tech-driven trajectory would lift real GDP by more than 2% over the baseline by 2033, or the equivalent of a US$500b boost over a decade.

Looking across major economies, the contributions from greater GenAI investment could also be significant. While the US market is likely to remain the leader in GenAI technologies investment, China and Europe will be following closely behind. We estimate that the lift to global GDP could total between US$300b and US$600b over the next five years. The boost to global potential GDP could amount to between US$500b and US$1t over the next decade.

In this installment of our “Economic impact of AI” series, we will focus on the business investment and capital accumulation dimension, and leave the productivity dimension of accelerating processes, optimizing operations and unlocking new capabilities to the next article in our series.

We will discuss investment in GenAI and associated capital accumulation by taking a deeper look at the following:

Businesswoman leading conference room meeting
1

Chapter 1

Back to basics: demand and supply

Capital investment in GenAI can spur stronger capital accumulation and productivity, boosting the global economy’s growth rate.

In an era where technological innovation is the cornerstone of economic prowess, GenAI has the potential to reshape the contours of businesses and the broader economy. This chapter delves into the burgeoning role of increased capital investment in AI, underscoring its potential to be a significant driver of near-term economic growth.

 

It’s important to consider that GenAI investment is not just a technological upgrade but a strategic economic lever to redefine business models, markets, industries and the very fabric of the global economy. By dissecting the dynamics of AI investment, we aim to unveil how it can propel economic activity, observed through the dual prisms of demand and supply.

 

From the demand perspective, investment in GenAI is seen as a new frontier for capital allocation, influencing various sectors from health care to finance, and energizing them with innovative capabilities. The investment fuels the industries it permeates, leading to an uptick in overall economic activity and consumer demand.

 

On the supply side, investment in AI will be a catalyst for stronger capital accumulation as well as productivity growth, lifting the global economy’s potential growth rate.

 

As we noted in the first installment of our series, prior general-purpose technologies have had a significant impact on economic activity, but that impact has generally lagged.

 

Some of the main reasons for that lag are implementation and diffusion delays, learning and adjustment periods due to the time it takes to effectively use new technologies and delays in the development of complementary innovations or infrastructure for the technology to be fully effective.

 

To establish GenAI as a cornerstone of modern industry, substantial capital investment may be required.

  • Research and development (R&D): Building and refining AI models necessitate a significant influx of resources. The data-intensive nature of GenAI calls for investment in gathering, storing and processing data, as well as in the computational power needed to train sophisticated models.
  • Infrastructure providers: Investment in the physical and digital infrastructure necessary to support AI technologies forms another cornerstone of this economic transformation. This encompasses everything from data centers to advanced networking capabilities and even cybersecurity. The adequacy of this infrastructure directly impacts the efficiency and effectiveness of AI solutions.
  • Software creation: The investment in AI applications across various business sectors is perhaps the most visible aspect of AI’s economic influence. From finance to manufacturing, AI applications are revolutionizing traditional business processes, enhancing customer experiences and opening new revenue streams. These investments are not merely about automating routine tasks but are also about leveraging AI to uncover insights, predict trends, and create more personalized and efficient services.
  • Corporate adoption: It’s essential for businesses to invest in integrating GenAI into their operations. This includes not only the technology itself but also the training of personnel and restructuring of processes to fully capitalize on AI’s potential. The widespread adoption of AI by businesses could have a notable economic impact as it leads to increased operational efficiencies, reduced costs and enhanced competitive capabilities. Moreover, as AI becomes more ingrained in business operations, it will likely drive the demand for skilled workers and AI-related services, and, consequently, it will probably stimulate job creation and economic activity in related sectors.
Diverse group of business people having a meeting in the conference room
2

Chapter 2

Demand perspective: contribution of capital in GenAI to GDP

Rising capital investment in GenAI is positioned to increase quickly and prompt GDP growth.

In assessing the potential economic impact of GenAI from a demand perspective, it is instructive to draw parallels with the investment dynamics of previous technological revolutions. In the early 1990s, business investment in information processing equipment and software totaled about 3% of US GDP, or US$155b.

As businesses invested in the physical and human infrastructure necessary to support, implement and reshape business processes in the computer age, that share of investment rapidly grew to 4.5% of US GDP, or US$400b by the early 2000s.


We are likely on the cusp of a similar trend with GenAI, where burgeoning investment in AI technology is poised to increase rapidly and boost GDP growth. Specifically, we isolated the following investment categories likely to capture AI technology:

  • Software
  • R&D in semiconductor and other electric components manufacturing, other computer and electronic product manufacturing, scientific services and software publishers
  • Computers and peripheral equipment
  • Communication equipment

Scenario analysis

The categories where new investment in GenAI will be most significantly captured totaled about US$750b in 2017, or about 3.8% of real US GDP. By 2022, investment had grown to just over US$1.1t, or about 5.2% of GDP.

  • Trend growth: Assuming trend growth in line with economic momentum from 2017 to 2022, investment would be expected to grow around 8.8% per year from 2023 to 2028 and represent 7.6% of real GDP by 2028, or US$1.9t. While this would mean that investment in AI technology would contribute about 0.4 ppt to GDP growth per year, it would not represent an increase in the growth contribution relative to recent past.
  • Baseline expectations: If, instead, we assume that nominal capital investment in AI technology grows 25% faster than the 2017-2022 trend pace over the next five years, then investment represents 8.1% of real GDP by 2028, or US$2t. This would translate into an incremental contribution of GenAI technology investment to GDP growth of 0.1 ppt per year (for a total contribution of 0.5 ppt), and by 2028, a boost to real GDP worth US$150b, or 0.6%.
  • Reason for optimism: Still, there may be reason to be even more confident about the outlook. Assuming capital investment in AI technology grows 50% faster than the 2017–2022 trend pace over the next five years — which is akin to the acceleration in business investment in information processing equipment and software in the late 1990s — then investment would grow about 11% annually from 2023 to 2028 and represent 8.7% of GDP by 2028, or US$2.1t. This would constitute an incremental short-term contribution to GDP growth of 0.2 ppt per year (for a total contribution of 0.6 ppt), and by 2028, a boost to real GDP worth US$300b, or 1.3%.

The potential uplift to global GDP from increased GenAI investment could also be substantial. With the US expected to continue leading in GenAI technology investment, closely followed by Europe, Japan and China, global GDP could see an augmentation of between US$300b (in our baseline scenario) and US$600b (in the optimistic case) over the next five years. This significant boost would reflect the accelerated adoption and integration of GenAI technologies across major economies, underlining the transformative impact of AI.

Smiling businesswoman at laptop in meeting in highrise office
3

Chapter 3

Supply perspective: strong capital foundation promotes growth

Past tech disruptions and our scenario analysis provide a case for optimism about GenAI’s ability to drive long-term growth.

At the heart of AI’s transformative potential on the supply side of the economy is its capacity to drive greater capital accumulation and stronger productivity growth. Capital investment in AI is not just an expenditure; it’s a strategic allocation of resources that acts as the foundation for developing and deploying AI solutions and seeds future productivity enhancements.

While we will delineate the long-term growth implication from GenAI-driven productivity growth in a subsequent chapter, we believe it is essential to dissect the impact of greater capital accumulation first.

Capital accumulation in AI involves investing in various components such as AI models (through building and refining), physical and digital infrastructure, software, AI applications, and AI integration and adoption. Just like physical capital, these investments in AI technologies act as the foundation that allows for stronger economic potential.

Capturing longer-term impact from greater capital investment in AI technology

The surge in business investment in information processing equipment and software through the 1990s did not just lead to a direct boost to GDP growth, but it also led to increased capital accumulation that then supported stronger long-term GDP growth.

To put things in perspective, the US economy’s potential GDP growth rate was estimated to be around 2.5% from 1990 to 1995, but subsequently it accelerated to 3.8% in the 1995-2000 period. Taking all drivers of growth into consideration, the capital contribution to potential GDP growth nearly doubled from 0.7 ppt in the early 1990s to 1.3 ppt in the 1995-2000 period. At the same time, the contribution of productivity also rose from 1.1 ppt to 1.7 ppt from 1995 to 2000 and remained elevated around 1.5 ppt from 2000 to 2005.

This confirms the findings from chapter 1, which indicated a five- to 10-year delay between the development of new technologies and their more sustainable impact on productivity and growth potential.


Scenario analysis

Using the same three scenarios, which analyzed the potential short-term economic impact of greater capital investment in AI technologies, we can infer the likely boost to potential US GDP growth in the five years from 2028 to 2033.

  • Trend growth: Assuming business investment in AI technology continues to grow in line with its moderate 2017-2022 trend, the annual capital contribution to long-term GDP growth in the 2028-2033 period would likely be around 0.5 ppt.
  • Baseline expectations: If, instead, we assume that capital investment in AI technology grows 25% faster than the 2017-2022 trend pace over the next five years, the capital contribution to long-term GDP growth in the 2028-2033 period would rise from 0.5 ppt annually to 0.6 ppt — thereby lifting the potential growth rate of the economy by 0.1 ppt. This may appear to be a small difference, but by lifting the economy’s potential growth rate, this stronger tech-driven trajectory would lift GDP by nearly 1% over the baseline by 2033, or the equivalent of a US$230b boost over a decade (US$360b in nominal terms).
  • Reason for optimism: As we noted earlier, there is reason to be more confident still about the potential capital accumulation contribution to long-term growth. Assuming capital investment in AI technology grows 50% faster than the 2017-2022 trend pace over the next five years – which is akin to the acceleration in business investment in information processing equipment and software in the late 1990s — the capital contribution to long-term GDP growth in the 2028-2033 period would rise from 0.5 ppt annually to 0.7 ppt — thereby lifting the potential growth rate of the economy by 0.2 ppt. This stronger tech-driven trajectory would lift real GDP by nearly 2% over the baseline by 2033, or the equivalent of a US$475b boost over a decade.

In the long run, the potential upside to global GDP from greater capital investment could be quite significant. How significant? Factoring stronger investment in Europe, Japan and China and slower investment across emerging markets, we estimate a boost to potential GDP growth worth between 0.5% and 1% by 2033, representing between US$500b and US$1t.


Five recommendations for business leaders

By focusing on the following areas, stakeholders can better navigate the complexities of AI capital investments and harness their full potential to drive meaningful business transformation.

Summary

This is the second installment of the EY-Parthenon macroeconomic article series on the economic impact of AI. The series aims to provide insights on the economic potential of GenAI, including new developments and actionable insights to arm companies’ decision-makers. The second article in this series covers the realm of capital investment in GenAI. Additional articles in the series will be published in the coming months.

Related article

Tech disruptions can inform the economic impact of AI

Discover learnings from three key historical episodes of rapid technological change that may help predict the economic impact of AI.

Harness the productivity potential of GenAI

Explore the likelihood of future substantial productivity gains from the use of generative AI across sectors. Learn more.

Five generative AI initiatives leaders should pursue now

Learn how to move beyond quick efficiency gains to a cohesive AI strategy that maximizes your growth potential in a fast-changing space.

    About this article