Artificial Intelligence and the US Economy: An Accounting Perspective on Investment and Production

Luisa Carpinelli Filippo Natoli Marco Taboga · 2026

Read Artificial Intelligence and the US Economy: An Accounting Perspective on Investment and Production by Luisa Carpinelli Filippo Natoli Marco Taboga (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2601-11196.txt.

Introduction

Luisa Carpinelli, Filippo Natoli, and Marco Taboga ask how AI shows up in the books. Their 2026 preprint looks at the United States economy. AI now sits at the core of talks on policy and markets. Yet its true size in the economy is still not well known.

The authors give a survey of how the new AI wave is logged. They check how it is counted in the national accounts of the United States. For this they join two tools in one frame. One is a plain accounting frame at the macro scale. The other is a simple sketch of how AI goods are made, step by step.

Catalog record

CREATORLuisa Carpinelli Filippo Natoli Marco Taboga
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE57 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2601-11196

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Artificial Intelligence and the US Economy: An Accounting Perspective on Investment and Production Thanks: The views expressed in this paper do not necessarily reflect those of the Bank of Italy or ESCB. The authors would like to thank Francesco Paolo Conteduca for sharing and support on the Trade Monitor Data. They also thank Riccardo Cristadoro for comments. Luisa Carpinelli Filippo Natoli Marco Taboga Abstract Artificial intelligence (AI) has moved to the center of policy, market, and academic debates, but its macroeconomic footprint is still only partly understood. This paper provides an overview on how the current AI wave is captured in US national accounts, combining a simple macro-accounting framework with a stylized description of the AI production process. We highlight the crucial role played by data centers, which constitute the backbone of the AI ecosystem and have attracted formidable investment in 2025, as they are indispensable for meeting the rapidly increasing worldwide demand for AI services. We document that the boom in IT and AI-related capital expenditure in the first three quarters of the year has given an outsized boost to aggregate demand, while its contribution to GDP growth is smaller once the high import content of AI hardware is netted out. Furthermore, simple calculations suggest that, at current utilization rates and pricing, the production of services originating in new AI data centers could contribute to GDP over the turn of the next quarters on a scale comparable to that of investment spending to date. Short reinvestment cycles and uncertainty about future AI demand – while not currently acting as a macroeconomic drag – can nevertheless fuel macroeconomic risks over the medium term. JEL: E2, O3, O4 Keywords: artificial intelligence, capital expenditures, data centers, national accounts 1 Introduction The rapid development and adoption of artificial intelligence (AI) technologies have been at the forefront of policy, market, and academic debates. From a financial stability perspective, a recurring question is whether the sharp appreciation in the share prices of technology companies is justified by a credible and sustainable expansion in their underlying profitability. From a more structural economic perspective, the key theme is how AI adoption will affect productivity and employment. More recently, a conjunctural issue attracting widespread attention from the media and market analysts is the extent to which AI‑related capital investment is currently contributing to US GDP dynamics. US growth in 2025 has been resilient above expectations, and some narratives suggest that investments primarily related to the rapid acceleration in the build‑up of data centers (Figure 1 ) have been the engine of this resilience. Figure 1 : Private construction spending by selected industries. Source. U.S. Census Bureau, Value of Construction Put in Place. Last observation: August 2025. This paper provides a tentative assessment of how AI production is being reflected in GDP components. We combine a macro-accounting perspective with a detailed description of the AI production ecosystem to answer a simple question: how, and through which channels, is the current AI wave showing up in measured US GDP? We focus on the parts of the AI value chain that are already large enough to matter for the short-term outlook – notably investment in data‑center infrastructure and the provision of cloud and AI services – and on the distinction between domestic value added and foreign content embedded in AI‑related spending. Assessing the incidence and channels of AI for the economic outlook is highly relevant for several reasons. For starters, it can guide policymaking, in particular monetary policy, by improving the understanding of inflationary pressures and labor market dynamics – for instance, by helping to reconcile the apparent dichotomy between resilient overall growth and a weak labor market. More broadly, a clear mapping of how AI investment and production enter GDP is essential for evaluating the strength and composition of demand, for gauging the domestic versus foreign share of value added in AI supply chains, and for informing fiscal and industrial policies aimed at supporting high-value activities. It also matters for financial stability analysis, as it helps distinguish between genuine, income-backed growth in AI-related sectors and developments that are mainly driven by asset price revaluations. Quantification is nevertheless fraught with methodological and measurement challenges. First, isolating AI-related activity in the national accounts is difficult because AI is typically embedded in broader aggregates – such as IT equipment, semiconductors, software, R&D, and IT services – and because it is often impossible to pin down the services where AI is incorporated. In practice, statistics frequently allow to capture only a broad “technology” sector rather than AI specifically. Second, firm-level AI figures are also rarely observable: key players in the AI production ecosystem are large, multi-activity firms whose financial statements combine costs, revenues, and profits across many lines of business. As a result, isolating the contribution of AI to costs, revenues and profits is extremely challenging. Third, the broader “second-order” effects set in motion by the AI transformation, including changes in productivity and employment linked to its adoption, are still too nascent to be reliably assessed. As such, they are outside the scope of this paper, which adopts a mechanical accounting perspective on how AI surfaces in GDP components. Three findings stand out. First, the boom in IT and AI‑related capital expenditure has provided exceptional support to aggregate demand, but its net contribution to US GDP growth is more limited than headline investment figures suggest because a large share of the associated hardware is imported. This reflects the fact that, while activities like chip design, data center operation, cloud services, and model development are mainly domestic, only part of the value created along the AI hardware supply chain accrues to the United States, as the fabrication and assembly of servers and accelerators are concentrated abroad. Second, data centers are clearly the backbone of the AI ecosystem. They concentrate the massive computational capacity required for AI model training, inference, and cloud‑based services, and translate AI‑related capital spending into productive output. From a macroeconomic standpoint, the construction and equipment procurement of data center facilities generate significant demand for construction services, high‑performance hardware and ancillary inputs, while the computational services they provide appear in the national accounts as final consumption and constitute intermediate inputs for other industries. Hence, data center activity is a pivotal channel through which AI investment feeds aggregate demand and contributes to GDP growth, making it a central focus of any macroeconomic analysis of the AI wave. Third, as the new generation of AI data centers becomes fully operational and sells computational and AI services to final users at home and abroad, the resulting revenue stream is expected to add to GDP at a level comparable to the capital expenditure needed to expand the underlying AI production capacity. This follows from the high utilisation rates of existing data center infrastructure, the current pricing of GPU‑based services, and the short pay‑back periods estimated in the paper. The rest of the paper is organized as follows. Section 2 provides a roadmap to the AI industry, describing the main actors in the ecosystem – hardware vendors, providers of cloud infrastructure and AI labs – and highlighting the central role of data centers and global supply chains. Section 3 turns to AI investment and quantifies how much of

Read the complete text