Artificial intelligence adoption and business performance among small and medium enterprises in Tanzania

Abstract

Artificial intelligence (AI) is increasingly applied in business operations, yet empirical evidence on its adoption and performance implications among small and medium enterprises (SMEs) in Tanzania remains limited. This study assessed the level of AI adoption and examined its relationship with business performance among Tanzanian SMEs. The study was anchored in the Technology–Organisation–Environment (TOE) framework and Resource-Based View (RBV). A quantitative cross-sectional survey design was employed. Data were collected using a structured, self-administered questionnaire from 320 SME owners and managers in Dar es Salaam, Arusha, Mwanza and Dodoma. AI adoption and business performance were measured using five-item reflective constructs on a five-point Likert scale. Data were analysed using partial least squares structural equation modelling (PLS-SEM) in SmartPLS 4. Results revealed a relatively high level of AI adoption (M = 3.45, SD = 0.82). AI adoption had a positive and statistically significant relationship with business performance (β = 0.533, p < 0.001), explaining 28.4% of its variance. The findings support the RBV perspective that AI can function as a valuable technological resource. SMEs should integrate AI into relevant operations while strengthening employee skills and data practices, while policymakers should improve training, digital infrastructure and affordable AI access.