Abstract Purpose. Federal statistics on business ownership, broad band connectivity, workforce skills and the adoption of artificial intelligence (AI) are generated and analyzed independently. This study asks whether the disadvantages they record co locate in the same firms, sectors and places, a pattern we call a “fracture line”, and which structural factors explain why AI adoption differs across U.S. states. Methodology. Based on the technology–organisation–environment (TOE) framework and the digital-divide theory, a state level dataset (50 states and the District of Columbia) is constructed using data from the Business Trends and Outlook Survey (BTOS) of the Census Bureau, the American Community Survey, the Annual Business Survey, County Business Patterns and BLS payroll data. The analysis has four parts: (i) inverse-variance-weighted tests of the national firm-size gradient from the BTOS AI Supplement (November 2025–February 2026); (ii) a shift-share decomposition separating industry-mix from within-industry effects; (iii) OLS models with HC3 robust errors and five robustness checks; and (iv) spatial diagnostics (Moran’s I and a spatial lag model). Findings. AI adoption is flat among firms with 1–19 employees (slope p = .35) and rises sharply from 20 employees (slope change = 4.47 points, p = .002). The gap separating micro firms from those with 250+ employees is 13.0 points in current use and 19.2 points in planned use. Across states, adoption ranges from 10.8% to 23.2%. Industry mix explains only 12 19% of that variation; the remaining four-fifths arises within industries. Household broadband subscription is the most robust correlate of adoption (about 1.7 points per standard deviation) in every specification and in an independent replication window. Ownership composition has no independent association, which supports a compositional rather than a constraint-based “double divide”. Five disadvantage indicators co-locate strongly (Cronbach’s α = .83), and their composite correlates −.55 with AI adoption. Originality/value. This is the first study to combine these federal programmes in a formal test of whether digital disadvantages co-locate. It shows that the AI divide among firms has a threshold rather than a linear shape, that it is mainly a within-industry phenomenon, and that among the measured place-level conditions it follows digital access most closely. Each of these findings points to a different policy lever.

REFERENCES  [3].

[1].Acemoglu, D., Anderson, G. W., Beede, D. N., Buffington, C., Childress, E. E., Dinlersoz, E., Foster, L. S., Goldschlag, N., Haltiwanger, J. C., Kroff, Z., Restrepo, P., & Zolas, N. (2022). Automation and the workforce: A firm-level view from the 2019 Annual Business Survey (NBER Working Paper No. 30659). National Bureau of Economic Research. https://doi.org/10.3386/w30659

[2].Anselin, L. (1988). Spatial econometrics: Methods and models. Kluwer Academic. https://doi.org/10.1007/978-94-015 7799-1

[3].Bloom, N., Sadun, R., & Van Reenen, J. (2012). Americans do IT better: US multinationals and the productivity miracle. American Economic Review, 102(1), 167–201. https://doi.org/10.1257/aer.102.1.167

[4].  Bonney, K., Breaux, C., Buffington, C., Dinlersoz, E., Foster, L., Goldschlag, N., Haltiwanger, J., Kroff, Z., & Savage, K. (2024). Tracking firm use of AI in real time: A snapshot from the Business Trends and Outlook Survey (NBER Working Paper No. 32319). National Bureau of Economic Research. https://doi.org/10.3386/w32319 .].

[5]Bonney, K., Breaux, C., Dinlersoz, E., Foster, L., Haltiwanger, J., & Pande, A. (2026). The microstructure of AI diffusion: Evidence from firms, business functions, and worker tasks (CES Working Paper No. 26-25). U.S. Census Bureau, Center for Economic Studies. https://www.census.gov/library/working-papers/2026/adrm/CES-WP-26-25.html

[6].Brynjolfsson, E., Rock, D., & Syverson, C. (2021). The productivity J-curve: How intangibles complement general purpose technologies. American Economic Journal: Macroeconomics, 13(1), 333–372. https://doi.org/10.1257/mac.20180386

[7].Bureau of Labor Statistics. (2026, September 18). State employment and unemployment — August 2026 (Table 3: Employees on nonfarm payrolls by state and selected industry sector, seasonally adjusted). U.S. Department of Labor. https://www.bls.gov/news.release/laus.t03.htm

[8]Chatterjee, S., Rana, N. P., Dwivedi, Y. K., & Baabdullah, A. M. (2021). Understanding AI adoption in manufacturing and production firms using an integrated TAM-TOE model. Technological Forecasting and Social Change, 170, 120880. https://doi.org/10.1016/j.techfore.2021.120880

[9]Dunn, E. S. (1960). A statistical and analytical technique for regional analysis. Papers of the Regional Science Association, 6(1), 97–112.

[10]. Fairlie, R. W., & Robb, A. M. (2008). Race and entrepreneurial success: Black-, Asian-, and White-owned businesses in the United States. MIT Press.

[11]. Forman, C., Goldfarb, A., & Greenstein, S. (2005). How did location affect adoption of the commercial Internet? Global village vs. urban leadership. Journal of Urban Economics, 58(3), 389–420. https://doi.org/10.1016/j.jue.2005.05.004

[12]. Government Accountability Office. (2026). Minority Business Development Agency: Information on performance assessment and compliance monitoring (GAO-26-107718). https://www.gao.gov/products/gao-26-107718

[13]. Grundy, A., Breaux, C., & Khatiwoda, D. (2026, May 26). Large firms with at least 20 employees biggest AI users. U.S. Census Bureau, America Counts. https://www.census.gov/library/stories/2026/05/ai-use-businesses.html

[14]. Hargittai, E. (2002). Second-level digital divide: Differences in people’s online skills. First Monday, 7(4). https://doi.org/10.5210/fm.v7i4.942

[15]. Keyhole Software. (2026). AI adoption rate by state (2026) [Compilation of U.S. Census Bureau BTOS state estimates, 18 May–9 August 2026]. https://keyholesoftware.com/ai-adoption-rate-by-state/

[16]. Kinkel, S., Baumgartner, M., & Cherubini, E. (2022). Prerequisites for the adoption of AI technologies in manufacturing – Evidence from a worldwide sample of manufacturing companies. Technovation, 110, 102375. https://doi.org/10.1016/j.technovation.2021.102375

[17]. Light Reading. (2026, September 3). BEAD ‘non-deployment’ funds to be used for deployment – NTIA. https://www.lightreading.com/broadband/bead-non-deployment-funds-to-be-used-for-deployment-ntia

[18]. Long, J. S., & Ervin, L. H. (2000). Using heteroscedasticity consistent standard errors in the linear regression model. The American Statistician, 54(3), 217–224. https://doi.org/10.1080/00031305.2000.10474549

[19]. McElheran, K., Li, J. F., Brynjolfsson, E., Kroff, Z., Dinlersoz, E., Foster, L., & Zolas, N. (2024). AI adoption in America: Who, what, and where. Journal of Economics & Management Strategy, 33(2), 375–415. https://doi.org/10.1111/jems.12576

[20]. Oliveira, T., & Martins, M. F. (2011). Literature review of information technology adoption models at firm level. Electronic Journal of Information Systems Evaluation, 14(1), 110–121.

[21]. Robinson, W. S. (1950). Ecological correlations and the behavior of individuals. American Sociological Review, 15(3), 351–357. https://doi.org/10.2307/2087176

[22]. Rogers, E. M. (2003). Diffusion of innovations (5th ed.). Free Press. [23]. Tornatzky, L. G., & Fleischer, M. (1990). The processes of technological innovation. Lexington Books.

[24]. U.S. Census Bureau. (2026a). Business Trends and Outlook Survey (BTOS), Artificial Intelligence Supplement, collected November 17, 2025–February 8, 2026 [Data files]. https://www.census.gov/hfp/btos/data

[25]. U.S. Census Bureau. (2026b). QuickFacts: United States and states [ACS 5-year 2020–2024; Annual Business Survey 2023; County Business Patterns 2023; 2020 Census]. Retrieved October 6, 2026, from https://www.census.gov/quickfacts/

[26]. U.S. Senate Committee on Commerce, Science, and Transportation. (2026, February 17). Cantwell, Moran reintroduce bill to help small business leverage AI tools [Press release]. https://www.commerce.senate.gov/2026/2/cantwell-moran reintroduce-bill-to-help-small-business-leverage-ai-tools

[27]. van Deursen, A. J. A. M., & van Dijk, J. A. G. M. (2019). The first-level digital divide shifts from inequalities in physical accessto inequalities in material access. New Media & Society, 21(2), 354–375. https://doi.org/10.1177/1461444818797082

[28]. Visual Capitalist. (2026). Mapped: The states where businesses use AI most [Compilation of U.S. Census Bureau BTOS state estimates, six releases, 15 January–26 March 2026]. https://www.visualcapitalist.com/mapped-ai-usage-by businesses-by-state-in-2026/

[29]. Whitacre, B., Gallardo, R., & Strover, S. (2014). Broadband’s contribution to economic growth in rural areas: Moving towards a causal relationship. Telecommunications Policy, 38(11), 1011–1023. https://doi.org/10.1016/j.telpol.2014.05.005