Artificial intelligence chatbot adoption in Mediterranean agrifood SMEs: A Q-methodology study

Artificial intelligence (AI) is rapidly reshaping how firms operate, yet small and medium-sized enterprises (SMEs) continue to lag in adoption, constrained by limited capital, scarce technical expertise, and resistance to change. In particular, how AI is perceived and adopted within Mediterranean agrifood SMEs remains under-explored, as research on this traditionally low-tech sector has so far privileged technical applications over managerial uptake. As accessible content-generation tools such as AI chatbots (e.g., ChatGPT, Gemini, Copilot) lower the barrier to entry, their uptake becomes not merely a technical matter but also an organizational and perceptual one. In this context, the present study aims to explore the subjective perspectives of Mediterranean agrifood SME leaders regarding the integration of AI chatbots into their daily operations. Theoretically, the study adopts the Technology-Organization-Environment (TOE) framework as its primary structure. The TOE was selected because it conceptualizes technology adoption as the joint outcome of a firm’s technological context, its internal organizational characteristics, and the external environment in which it operates. This integrated lens is particularly suited to the resource-constrained, tradition-bound, and institutionally heterogeneous setting of Mediterranean agrifood SMEs, where adoption cannot be fully explained by technical or economic factors alone. Methodologically, it employs Q methodology, a quali-quantitative approach to human subjectivity. A 36-item Q-set built on a TOE design matrix, in which each of the three dimensions (technological, organizational, and environmental) was articulated into four thematic subgroups of three statements each, spanning perceived drivers and barriers. A purposive sample of 22 agrifood business owners and managers familiar with AI chatbots, drawn from 13 Mediterranean-bordering countries, sorted the statements along a forced quasi-normal grid (−5 to +5) under the common stem “In my opinion, AI chatbots in my business…”. Post-sort questions and a socio-demographic survey support the final interpretation. Preliminary analysis suggests four distinct leader perspectives, namely (i) Regulatory innovators, (ii) Pragmatic efficiency adopters, (iii) Governance skeptics, and (iv) Context constrained adopters. Adoption is expected to hinge less on the usefulness of AI chatbots than on what legitimizes them in each leader’s eyes, reflecting institutional, organizational, and infrastructural conditions rather than technical or economic merit alone. This would carry context-sensitive implications for policymakers and SME leaders pursuing digital transformation.