FATE in AI: Towards Algorithmic Inclusivity and Accessibility

Abstract

With Artificial Intelligence (AI) occupying the centre stage of technological advancements, its impact is affecting many sections of society. Because algorithmic decisions carry both economic and personal implications, fairness, accountability, transparency and ethics (FATE) in AI are geared towards checkmating algorithmic disparities. However, one of the noted limitations of the present discourse on such critical issues is the dominance of the more economically developed countries (MEDC), resulting in neglecting local knowledge, cultural pluralism and global fairness. This study builds upon existing research on responsible AI, with a focus on areas in the Global South deemed under-served vis-a-vis AI. Our goal is two-fold (1) to assess FATE-related desiderata with emphasis on transparency and ethics and (2) to offer insights and proffer recommendations to stimulate action towards representative and responsible AI. We designed a user study (n = 43) and a participatory session (n = 30) to achieve the above goals. Our findings reveal a path towards encoding bias and amplifying a form of stereotype by AI models. For improved inclusivity, we propose a community-led strategy to operationalise the collection and curation of representative data towards responsible AI design. The initiative will empower the affected community or individuals to probe and police the growing application of AI-powered systems. Moreover, we offer some recommendations, informed by the public, that adhere to social values as core requirements for practitioners to incorporate context-specific FATE in AI needs.

Ähnliche Arbeiten