Artificial Intelligence-Driven Digital Financial Inclusion and Household Financial Resilience: Evidence from Sierra Leone

Abstract

The ability of AI-powered financial technologies to reach, afford and improve the quality of formal financial services is
increasingly a key determinant of household financial resilience in low-income economies. The reach, affordability and quality of
formal financial services is increasingly a driver of household financial resilience in low-income economies, and artificial
intelligence (AI)-enabled financial technologies are playing an increasing role in achieving this. While there is a growing number
of scholarships that connect digital financial inclusion to welfare outcomes for households, there is comparatively little empirical
evidence on how AI-powered digital financial services can contribute to resilience of fragile, low-income economies like Sierra
Leone. This study addresses this gap by investigating the mediation role of digital financial inclusion and financial capability
between AI-driven digital financial services and financial resilience of households, and the moderating role of digital trust between
these two relationships. Primary data was collected using a cross-sectional survey design with 400 respondents from urban and
rural districts of Sierra Leone and analyzed with reliability and validity diagnostics, hierarchical multiple regression and
bootstrapped mediation analysis with 5,000 resamples. The measurement model showed good internal consistency and discriminant
validity, and common method bias was not considered to be a serious threat. The results indicate that AI-enabled digital financial
services have a significant predicting power for digital financial inclusion, which in turn has a significant predicting power for the
financial resilience of households both directly and indirectly through financial capability, in line with a serial mediation pathway.
Financial capability and financial inclusion were included to assess the full mediation of the direct effect of AI-driven digital
financial services on resilience, and the result was statistically non-significant. The direct effect of digital trust was significant, but
there was no significant moderation of the pathway between AI-driven-services-to-inclusion. The results contribute to the theory
of digital financial inclusion in the AI era and provide concrete policy recommendations for regulators, fintech companies and
development partners to enhance household resilience in Sierra Leone and other similar Sub-Saharan African settings.