ABSTRACT: This article analyzes the contribution of artificial intelligence (AI) to the performance of small and medium-sized enterprises (SMEs) in emerging economies. Methodologically, the research adopts a mixed approach. The quantitative survey, conducted among 353 SMEs (trade, agribusiness, services, technologies), in Central Africa, is estimated by structural equations (PLS-SEM). The results indicate that AI exerts a significant direct effect on performance (β = 0.29; p < 0.001) and an indirect effect via serial mediation AI → inclusive innovation → resilience → performance (β_indirect ≈ 0.04; p < 0.05). Inclusive innovation predicts resilience (β = 0.41; p < 0.001), and resilience predicts performance (β = 0.28; p = 0.011). The model has substantial explanatory power (R² ≈ 0.48). The thirty-five qualitative interviews confirm these mechanisms, while highlighting the obstacles: the cost of entering solutions, the lack of digital skills, institutional instability.

On a theoretical level, the article shows that AI becomes performative if, and only if, it is socialized by inclusive innovation and consolidated by resilience. As a result, public policies must go beyond the diffusion of technology to strengthen the accessibility and absorption capacity of SMEs.