Fed paper reveals the reality gap behind the AI boom: the productivity dividend has yet to materialize, inflation costs are emerging but limited in extent, and employment effects are difficult to quantify.

According to a research paper released by Federal Reserve staff, AI demand has driven a surge in U.S. data center investment and raised the prices of some tech inputs, but its impact on overall productivity and employment remains limited. This means that the current AI boom's boost to the U.S. economy is mainly reflected in investment growth brought by data center construction and purchases of chips and related equipment, while whether AI technology can truly broad enterprise efficiency gains and a productivity dividend still needs to be tested over time. It is reported that the research, titled "Artificial Intelligence and the U.S. Economy: Current Evidence," was written by Nick Carollo, Leland Crane, Daniel Garcia Molina, and Joseph Nichols, and was published in November. The authors noted that although AI is very likely to have a huge and lasting productivity effect in the long run, only a few years have passed since ChatGPT was launched, and technology diffusion, corporate organizational adjustments, and the realization of economic effects remain at an early stage. It should be noted that this is a discussion paper by Federal Reserve staff, and its analysis and conclusions represent the authors' views and do not imply endorsement by the Federal Reserve Board or other staff. Data center investment plans exceed US$1 trillion, but import spending offsets part of the economic boost The paper found that investment plans for new data centers in 2025 exceeded US$1 trillion, up from nearly US$300 billion in 2024. Based on projects announced as of May 2026, under a central scenario, the annualized scale of data center-related investment is expected to exceed US$880 billion by the end of 2027, equivalent to more than 2% of GDP. Morgan Stanley estimates show that cash capital expenditure by hyperscalers and neoclouds is expected to continue growing, rising from US$946 billion in 2026 to US$1.66 trillion in 2027, and further reaching US$1.77 trillion in 2028. Alphabet (GOOGL.US) and Amazon.com, Inc. (AMZN.US) account for the largest shares of projected spending. The two companies' combined cash capital expenditure is expected to reach US$730 billion in 2027 and about US$765 billion in 2028. However, the scale of data center investment does not mean an equivalent increase in domestic economic output. The authors estimate that about half of related spending goes to imported goods, especially high-tech equipment, so the increase in imports will offset part of the boost to U.S. domestic output from investment. At the same time, data center construction may also drive supporting investment such as power infrastructure, but rising prices of high-tech equipment may also squeeze other companies' investment budgets. There is also uncertainty about projects actually being implemented. Data center construction requires land, power, water, labor, and large amounts of equipment, and insufficient supply of any one resource could push up construction costs and delay project timelines. Some local governments have already restricted data center projects due to electricity and water demand, and grid connection queues, shortages of natural gas turbines, and chip supply shortages also pose potential bottlenecks. The paper noted that delivery lead times for some natural gas turbines have been extended from two years to more than five years. The AI construction boom pushes up prices of some tech products, with an inflation impact worth noting but limited in extent The data center construction boom has also pushed up prices of products such as memory chips, data storage devices, bare circuit boards, and assembled circuit boards. In the 12 months through March 2026, prices for a category of personal consumption expenditures (PCE) covering video, audio, photographic, and information processing equipment and media rose nearly 8%; in the previous three years, prices for this category fell by about 3% per year on average. The paper noted that rising prices for software and accessories were also an important factor in the higher prices for this category, and some software companies may be passing on the cost of new AI features to users. Rising prices for data storage devices also contributed to related consumer prices. However, the authors believe that so far the impact of these price increases on the overall PCE price index has been significant but limited, and the impact could expand further in the future. Related research also found that since January 2025, inflation in the core PCE price categories most exposed to data centers has been about 5 percentage points higher than in categories with median exposure. This means that AI infrastructure construction may, in the short term, push up some costs through chips, electronic equipment, software, and electricity, while the cost-reducing effect of productivity gains has not yet fully appeared. Federal Reserve researchers believe that if AI can significantly raise productivity in the future, it could indeed ease inflation by lowering companies' production costs, but this effect cannot yet be observed. AI adoption continues to expand, but depth of enterprise use remains insufficient The scope of AI use among consumers and businesses continues to expand, but adoption remains uneven. A 2025 survey found that about 50% of adults use AI outside work, mainly for information gathering and writing; multiple surveys as of early 2026 showed that more than 40% of workers use AI at work. However, only about 10% of workers use AI at work every day, indicating that the technology has not yet become a core part of most employees' daily workflows. Survey results at the enterprise level vary widely. The U.S. Census Bureau's Business Trends and Outlook Survey shows that about 20% of businesses adopt AI; other studies put the proportion at 58% to 78%. This difference is related to the specific wording of survey questions, sample scope, and statistical weighting. Overall, larger businesses and younger firms have higher AI adoption rates, and industries such as Information Services Group, Inc., financial activities, and professional and business services are also relatively ahead in application. However, companies currently often use AI only in a few business functions, such as IT, human resources, marketing, finance, and strategic management, rather than fully integrating it across the entire organization. Particularly noteworthy is that AI adoption in R&D remains relatively low, even though R&D may be an important channel through which AI drives long-term innovation and productivity growth. Federal Reserve researchers believe that to achieve sustained and broad productivity gains, companies may need to further incorporate AI into core production processes, redesign organizational structures, automate more work tasks, and invest in supporting resources such as training and management. Simply purchasing AI tools or giving employees access is not enough to guarantee a large increase in productivity. The impact of AI on the job market has begun to appear, but the overall effect remains difficult to quantify AI's impact on the labor market remains mixed. A study cited in the paper found that after the launch of large language models, employment among workers aged 22 to 25 in occupations highly exposed to AI fell by 16% compared with workers of other ages in the same occupations and with workers in occupations less exposed to AI. Other studies have reached similar conclusions. The authors noted that this impact is mainly reflected in reduced hiring demand rather than large-scale layoffs. Workers entering the workforce may be more vulnerable because their work often includes more simple, standardized, and easily codified tasks, while experienced employees often possess specialized knowledge and firm-specific experience that is harder for AI to replace. At the same time, AI may also lead to job reassignment. Some companies prefer to retrain affected employees rather than lay them off directly. A survey by the New York Fed found that when facing roles potentially affected by AI, companies reported a higher proportion of employee retraining than layoffs. Therefore, AI's impact on overall employment and unemployment remains difficult to identify. The occupational groups currently more affected account for a relatively limited share of total U.S. employment, but as the scope and depth of enterprise AI adoption continue to expand, this impact may become more apparent in the future.