AI Enablement in SMEs: Why Theoretical Models Miss the Real Picture

Abstract

Artificial intelligence is widely promoted as transformative for business competitiveness, yet adoption in small and medium-sized enterprises (SMEs) remains limited and poorly understood. While large corporations develop formal AI strategies backed by dedicated teams and budgets, micro and small enterprises face a different reality: constrained resources, fragmented knowledge and the absence of internal AI expertise. Most existing theoretical frameworks do not account for these differences, as they derive from large-enterprise contexts and assume conditions that rarely exist in smaller firms. This study investigates how AI enablement unfolds in the SME context and where theoretical predictions diverge from empirical evidence. The research employs an exploratory sequential mixed-methods design (QUAL→QUAN). This paper presents findings from the qualitative phase, based on reflexive thematic analysis of in-depth interviews with nine experts who implement AI solutions in SMEs as consultants, integrators or advisers, coded in two cycles and validated through member checks. The interviews were conducted between 2022 and 2025. The findings challenge several established assumptions. Strategy in SMEs is not a formal document but a dynamic process that emerges through collaboration with external IT consultants and incremental implementation. The primary barrier is not technology, which has become increasingly accessible, but knowledge gaps among SME managers about what AI can and cannot do. Innovation culture, treated in most frameworks as a distinct dimension, functions in SMEs as part of human-resource dynamics. Most notably, partnership with external IT consultants emerges as the central mechanism, operating through a three-tier model of consulting, joint planning and long-term support. Dynamic capabilities in SMEs are therefore realised relationally, through external partnerships, rather than through internal organisational processes. The paper contributes a partnership-based refinement of dynamic capabilities theory for smaller firms, and argues that supporting AI adoption in such firms requires partnership-oriented management and policy rather than technology-centric strategies that overlook how these enterprises actually function.