JOURNAL: Augmentation, automation, or reconfiguration? A firm-level typology of artificial intelligence and workforce change in Slovakia

Whether artificial intelligence (AI) destroys or creates jobs is still contested, and most firm-level evidence comes from advanced economies and treats employment as a single net quantity, which hides what happens inside individual firms. This study asks how AI relates to workforce change among AI-engaged firms in Slovakia, a small open economy in Central and Eastern Europe. The data come from an organization-level survey of 816 firms; the 693 firms that use, are piloting, or plan to adopt AI answered two workforce questions. The questions ask whether the firm has introduced, or plans to introduce, new positions because of AI, and whether AI has led, or is expected to lead, to layoffs. Cross-classifying the two yes/no answers yields a four-way typology (Static, Augmenter, Reducer, Reconfigurer), analyzed with logistic, multinomial logistic, and bivariate probit models. Among AI-engaged firms, more report job creation than report layoffs (18.0% versus 9.2%; McNemar p < 0.001). The two outcomes also co-occur far more often than chance (odds ratio 7.23; bivariate probit error correlation ρ = 0.54), a pattern consistent with workforce reconfiguration rather than simple substitution, although shared-respondent reporting may contribute to it. The stage of AI adoption is associated with both outcomes, while the direction of change is associated with organizational orientation: AI enthusiasm is linked to fewer layoffs, and perceived employee job-threat concern to more. What a firm has deployed matters more than how mature it judges itself to be. The study shifts attention from net headcount toward the firm-level structure of job creation and reduction.

Authors: Peter Štetka, Zuzana Hajduová, Nora Grisáková