Creating Resilience
AI’s Impact on the Economy, Employment & Productivity
AI & Commercial Real Estate, Part 1 of 4
April 30, 2026 7 Minute Read
Executive Summary
This is the first in a four-part series of reports by CBRE Research on the impact of artificial intelligence on commercial real estate. This initial report examines how AI is already intersecting with core macroeconomic fundamentals and where its influence will likely be most consequential over the coming cycle.
- AI investment is rewriting the economic playbook in real time. Hyperscalers are projected to spend $3.7 trillion on AI infrastructure over the next five years—a buildout that experts estimate will eclipse the railroad expansion of the 1850s in relative scale. 2025 AI investment was equivalent to roughly half of U.S. GDP growth for the year.
- AI's potential impact on the labor market is nuanced. While nearly 80% of U.S. companies are using AI, close to 90% report little impact on their business so far. However, since 2022, employment among entry-level workers in the most AI-exposed occupations has been 13% lower than for workers in less-exposed roles.
- AI is arriving at a critical demographic inflection point. With U.S. working-age population growth slowing, AI-driven productivity gains could allow a stagnant labor force to generate comparatively strong GDP growth. Productivity growth is forecast to accelerate to 2.1% from 1.4% annually through 2030, with AI becoming the dominant driver of output as annual employment growth slows to just 0.5%.
- Job composition will change more than job count. Research shows a 60% correlation between an occupation's automation risk and its adaptive capacity, meaning roles will more likely be reconfigured than replaced. Just as 20% of today's jobs didn't exist in 1999, AI will create new occupations even as it replaces existing roles.
- Our next report in this series will explore the impacts of AI on office demand. Historically, office-using jobs have grown fastest following major technological advancements like the internet and smartphone. We expect the same to eventually occur in the AI era, albeit to a lesser degree.
A History of Technological Innovation
Enjoying this? Sign up to follow the series.
Even though it looks like overbuilding, investing early in transformative infrastructure pays off. As was the case with the internet, the same logic applies to AI today.
AI is no longer theoretical; today it is a reality and a major economic force. Early debate on the potential impact of AI focused on job displacement and headline risks. Nevertheless, history suggests a more powerful, enduring impact: AI can reshape the fundamental drivers of economic performance.
Like every prior wave of transformative innovation, this AI era also comes with a spectrum of prognosticators: true believers, doomsayers and skeptics. The early internet era was no different. Optimists promised it would usher in a golden age of productivity and convenience; pessimists warned it would eliminate white-collar jobs; and skeptics dismissed it as a wildly overvalued fad destined to end in a painful correction.
History showed that all three were right over time. Valuations proved frothy and the dot-com bubble burst in 2000. U.S. companies had overbuilt the information infrastructure, laying vast amounts of “dark fiber” well ahead of demand. But this eventually paid off, allowing the U.S. to harness the internet at scale sooner than other nations, fueling decades of productivity gains and helping underpin sustained U.S. economic outperformance.
Figure 1: Productivity CAGR (Output Per Worker)
Source: CBRE Research, CBRE Econometric Advisors, Oxford Economics, Q4 2025.
As the saying goes, “history does not repeat itself, but it often rhymes.” Today, AI is following a similar pattern to the dot-com era: a new technology that is impossible to price precisely given its unknown potential; a massive infrastructure buildout, particularly among U.S. companies; and different forecasts that range from the displacement of millions of white-collar workers to a more utopian outcome. Skeptics are once again raising bubble concerns.
Current Economic Impact of AI
AI infrastructure investment is a net positive for baseline economic growth, but carries both downside risk if valuations outpace revenue and meaningful upside if productivity gains occur sooner than expected.
Realizing the capability of AI as a driver of economic growth requires enormous investment. U.S.-based AI startups have raised roughly $578 billion in venture capital since 2020, with nearly three-quarters of that in the past two years, according to CBRE Research and Pitchbook. Since 2020, approximately 80% of this funding has gone to companies headquartered in the San Francisco Bay Area. This is having positive effects on local employment and office space demand. Investment is expected to remain elevated for years, supported further by a growing pipeline of AI-related IPOs.
Figure 2: U.S. AI VC Funding & Share of Total VC Funding
At the same time, large technology firms—known as hyperscalers—are making unprecedented capital investments to build the infrastructure that powers AI. CBRE estimates that hyperscalers spent approximately $1.1 trillion on capital expenditures between 2021 and 2025, largely for data centers, chips and computing capacity. Bloomberg predicts this capex to reach $3.7 trillion over the next five years, or roughly $740 billion annually. Many experts estimate that this digital bonanza will eclipse the railroad expansion of the 1850s in its scale relative to the U.S. economy.
Total investment in information processing equipment and intellectual property (effectively the AI buildout) last year was equivalent to nearly half of all U.S. GDP growth, compared with just 8% in 2023 and 2024, according to the U.S. Bureau of Economic Analysis. This level of investment is expected to continue this year. However, exposure to this speculative sector presents both upside and downside risks to medium- to long-term economic growth. CBRE expects 2.2% annual GDP growth through the end of this decade. On the downside, a disconnect between AI-driven valuations and actual revenue could potentially shave 105 basis points (bps) off annualized GDP growth over the next two years. Conversely, on the upside, a larger- and earlier-than-expected boost to productivity from AI investment would add 95 bps to annual GDP growth.
Although uncertainty remains high, one thing is clear: As AI adoption scales across industries, its effects on productivity, labor, capital allocation and space demand will be significant.
Figure 3: Rolling Two-Quarter GDP Growth: AI & Non-Tech Investment
Source: CBRE Econometric Advisors, U.S. Bureau of Economic Analysis, Q4 2025.
Total investment in information processing equipment and intellectual property (effectively the AI buildout) last year was equivalent to nearly half of all U.S. GDP growth.
AI’s Impact on Labor
While nearly 80% of U.S. companies are using AI, most report little impact so far. Yet early signs of disruption at the margins, particularly for entry-level roles, suggest the story is just beginning to unfold.
The economic impact and integration of AI into everyday business workflows and therefore productivity remains limited. According to the National Bureau of Economic Research, as of February 2026 nearly 80% of U.S. companies reported using AI, primarily for text generation. However, close to 90% said that while AI has had little impact on their business so far, they expect it will have a much larger effect over the next three years by boosting productivity and reducing labor demand.
The rise in unemployment among recent college graduates could suggest early impacts of AI, particularly on entry-level white-collar roles. However, unemployment among this cohort has been rising since at least 2018, well before the current wave of generative AI. Therefore, attributing recent labor-market softening primarily to AI overlooks a broader set of structural and cyclical factors. These include the current slow-to-hire, slow-to-fire environment as 15+ years of economic growth collides with elevated interest rates, heightened geopolitical risk and AI-driven questions around the future of work. The rise in unemployment across all sectors since 2021 is not being driven exclusively by AI, since this new technology was not available at scale until late 2022.
Figure 4: Unemployment Rate by Cohort
Source: U.S. Census Bureau, U.S. Bureau of Labor Statistics, Q4 2025.
Some emerging evidence suggests that AI is beginning to affect employment at the margin. Executive search firm Challenger, Gray & Christmas cites AI as a factor in 5.5% of U.S. layoffs last year, although some observers speculate that much of this is simply due to cost-management decisions. In addition, the National Bureau of Economic Research found that since 2022, employment among entry-level workers in the most AI-exposed occupations has been 13% lower than for workers in less-exposed roles and more senior workers in those same occupations.
These findings suggest that while AI is unlikely to be the sole or primary driver of recent labor-market trends, its effects are starting to surface. For commercial real estate, the key question is less whether total employment rises or falls and more how AI reshapes job mix and where high-value work concentrates. This turn will have direct implications for office utilization, including absorption levels and demand for flexible space.
For commercial real estate, the key question is less whether total employment rises or falls and more how AI reshapes job mix and where high-value work concentrates.
Productivity Potential of AI
With U.S. working-age population growth slowing, AI's productivity boost is arriving at the right time by allowing a stagnant labor force to punch well above its weight in driving GDP growth.
History offers useful guidance on how major technologies ultimately affect employment and economic growth. New technologies initially go through a long ramp-up period in which their economic impact is difficult to detect. It is only once they reach broad, global adoption that productivity accelerates and output rises meaningfully. In these global-adoption phases of technological innovations, office-using industries tend to show outsized employment gains. But at the same time, we see the spread between GDP and office-using employment widening each cycle in line with productivity gains. We therefore expect the recent outperformance of GDP relative to employment to accelerate as AI-driven productivity filters through the economy.
Job growth is contingent on the size of the working-age population. Across successive cycles of expansion and contraction, long-run employment has tracked labor force growth with remarkable consistency, shaped less by economic conditions than by demographic ones.
This constraint is especially relevant today as projections show the U.S. working-age population is flattening. CBRE projects U.S. population growth of 0.34% annually over the next decade, with the working age population expected to decline by 0.28% annually during that time. Both of these would be the slowest such growth rates in U.S. history. In that context, AI’s productivity potential may be arriving at an opportune time, allowing modest employment growth to translate into comparatively strong GDP growth.
CBRE expects U.S. annual GDP growth to average 2.2% through the end of this decade. With employment growth projected at just 0.5% per year over this period, productivity becomes the dominant driver of output growth in this outlook.
Figure 5: Real GDP, Employment & Working Age Population Indexed to 1990
With employment growth projected at just 0.5% per year over this period, productivity becomes the dominant driver of output growth in this outlook.
Productivity—perhaps the most closely scrutinized variable in the debate about AI’s potential impact—is forecast to grow at an annualized rate of 2.1% through 2030, a meaningful improvement from the 1.4% average annual growth over the past 25 years. This anticipated acceleration is attributable in large part to the adoption of AI across the economy.
There currently is little empirical evidence that rising productivity will exert downward pressure on overall employment growth. The biggest constraint on hiring is not labor demand, but the size and growth of the labor force itself. In this context, productivity gains that enable greater output from a largely fixed pool of workers likely will be economically beneficial and increasingly necessary given today’s demographic realities.
Figure 6: Annualized Productivity Gains by Decade
How many workers will be replaced by AI?
Just as the internet created more jobs than it destroyed—20% of today’s jobs did not exist in 1999—AI is poised to reconfigure rather than replace work. The most disrupted industries and markets ultimately stand to gain the most.
Studies show that the internet boom dramatically increased the pace of job displacement within existing industries but also accelerated the creation of new occupations and industries. Today, 20% of all jobs did not exist in 1999.
In other words, past periods of heightened job losses due to technological advancements were offset by new, more productive jobs and industries. AI appears poised to follow a similar path—this time against the backdrop of a labor force that is no longer growing.
Our research suggests that AI-driven changes to the composition of jobs will be far more impactful than changes to the level of employment. A recent study by GovAI fellow Sam Manning and Tomas Aguirre found a 60% correlation between an occupation’s automation risk and adaptive capacity (the ability of displaced workers to quickly adapt to new forms of work and regain employment). This reinforces our view that many of those most exposed to AI technology have skillsets that position them to see net gains in job opportunities. In other words, the greatest upside will emerge in industries and occupations most exposed to AI-induced disruption.
Figure 7: AI Automation Risk vs. Adaptive Capacity by Occupation
Source: CBRE Research, Sam Manning & Tomås Aguirre, Federal Reserve Bank of St. Louis, U.S. Census Bureau, Q4 2025.
Past periods of heightened job losses due to technological advancements were offset by new, more productive jobs and industries. AI appears poised to follow a similar path—this time against the backdrop of a labor force that is no longer growing.
Roles like web designers, software engineers and financial analysts are expected to see big shifts in the nature of their work but have the skills necessary to excel in an AI-integrated labor market. This isn’t to suggest that there won’t be any headwinds for the labor market as result of these changes, but those industries facing the most AI disruption will likely ultimately see the largest job gains. For example, despite being in the crosshairs of AI displacement, software development job postings were up by 15% since last year.
Jobs that have grown fastest since 2000 are also the most exposed to AI disruption because they are partly task-oriented, making them more adaptable to AI applications. However, as AI automates certain tasks, it will allow workers in these roles to focus more on human judgment, coordination and proficiency. As a result, these jobs will more likely be reconfigured rather than replaced, allowing productivity to rise without broad-based job losses.
Companies are looking for ways to apply AI for greater productivity at lower costs. This will put downward pressure on the labor market in the near term. Over the medium-term, job creation will likely outpace job replacement, eventually leading to both job growth and productivity gains. This productivity boost is coming at an opportune time given expectations of stagnant population growth in the years ahead.
Our next report in this series will explore the impacts of AI on office demand. Historically, office-using jobs have grown fastest following major technological advancements like the internet and smartphone. We expect the same to eventually occur in the AI era, albeit to a lesser degree.
Related Insights
-
AI is reshaping commercial real estate. Data scientist Sandy Pentland and CBRE's Sandeep Davé explore how emerging technology and human insight are driving smarter decisions for investors, occupiers, and building operators.
-
Artificial intelligence (AI) companies are driving strong demand for data centers across North America that have exceptionally large power capacity to run their applications.
-
Article | Evolving Workforces
Business Insights | Everyone Can Have a (Digital) Assistant, with Ellis AI
February 13, 2025
CBRE's Ellis AI platform revolutionizes commercial real estate with GenAI, enhancing productivity and efficiency through advanced AI digital assistants.
