MONOGRAPHY: The Diffusion of Artificial Intelligence in Slovak Enterprises Forecasting Adoption and Its Trajectory
This monograph turns from the description and comparison of artificial intelligence (AI) adoption toward its prediction. Building on prior evidence positioning Slovakia within the international benchmark landscape and assessing AI adoption among Slovak enterprises through dedicated enterprise survey waves, it models the diffusion of AI across Slovak enterprises and derives forward-looking scenarios of its future trajectory. The monograph deploys two complementary modelling approaches. The first is the Bass diffusion model, applied both to the sample as a whole and to eight functional areas of AI use, which characterises the shape and timing of the adoption curve and the mechanism driving it.
The second is a discrete-time Markov chain over three adoption states — non-adopter, planner and adopter, which makes explicit the transitions between these states and, in particular, the role of the planner pool that the binary logic of the Bass model necessarily abstracts away. The modelling is preceded by a descriptive analysis of the sample and followed by a dedicated robustness assessment, so that the forecasts are presented together with a transparent account of their sensitivity to the underlying assumptions.
The monograph complements empirical and comparative evidence on AI adoption with a predictive perspective. Its findings are addressed to researchers, policymakers and enterprise decision-makers seeking to understand not only where AI adoption stands today, but how quickly, in which functional domains, and through which behavioural transitions it is likely to spread in the coming years. Consistent with the methodological caution that characterises this study as a whole, the forecasts presented here are explicitly conditional on a defined set of assumptions, which are stated transparently and revisited in the discussion of limitations and in the robustness analysis. They should be read as structured, model-based scenarios rather than as unconditional predictions.
Authors:
Vladimír Hojdík, Andriana Baziuk, Simona Kosztanko
