Toward Financial Inclusion

Language’s centrality in technology, or the expression of Artificial Intelligence (AI) through language, should be evaluated to consider whether the Fourth Industrial Revolution (4IR) technologies in banking can be harnessed for equity or whether they risk perpetuating digital divides. A subset of Human Language Technologies (HLT) and Virtual Assistants (VAs) are conversational applications that use machine learning – the use of computers to draw patterns and inferences from large sets of data to provide human-like interactions to customers or users of the software on mobile apps or websites. While assertions that 4IR and its proponents (AI, machine learning, robotics) will have positive and radical impacts, it is crucial to question whether the Virtual Assistants (VAs) used in financial technologies (Fintech) will enable financial access by bridging language barriers rather than risking current forms of exclusion, disparity, and digital gaps. The aim of this paper is to evaluate whether the expression of language in Fintech can be harnessed to promote equity or worsen the digital divide. This is foregrounded by a Critical Discourse Analysis theory to highlight how language manifests in Fintech. Employing an interpretive qualitative methodology, data were collected through semi-structured interviews and focus groups to unearth rich, contextual insights. Thematic analysis was used to unravel imaginaries surrounding VAs and to extract meaningful patterns. This paper demonstrates that, while great in theory, accommodating the rich diversity of languages spoken in South Africa often runs counter to the pressures of the market in practice. Therefore, this study looks to catalyse policy development that will ensure that HLTs, like VAs, incorporate indigenous languages. This can be achieved by subsidising the development of Free and Open-Source solutions for use in the financial sector. These policies and solutions can be developed in collaboration with local communities, who can provide valuable feedback on their experiences with VAs.