Communicative Transparency by Design: a conceptual framework for explainable AI as infrastructure for accessible digital communication

Abstract

The opacity of algorithmic systems embedded in digital platforms constitutes a structural barrier to communicative participation and digital inclusion. While Explainable Artificial Intelligence (XAI) has been extensively theorised as a technical and cognitive problem, its communicative implications remain undertheorised within Communication Studies. This article proposes the Communicative Transparency by Design (CTbD) framework, which reconceptualises XAI mechanisms as communicative infrastructure governing how platforms explain their decisions to heterogeneous publics. Synthesising platform studies, human-centred XAI, and algorithmic-literacy and accessibility research through an integrative conceptual analysis, the framework articulates three interdependent dimensions: Inferential Accessibility (the intelligibility of explanations for diverse user profiles), Participatory Agency (users’ capacity to contest decisions through communicative acts), and Equitable Intelligibility (the differential distribution of transparency benefits across literacy, disability, and socioeconomic status). The framework’s specific novelty is to shift the unit of analysis from the accuracy of an explanation to the communicative relationship it establishes: where prior accounts of algorithmic transparency and accountability ask whether a system discloses its reasoning, CTbD asks to whom that disclosure is addressed, with what affordances for response, and with what equity of effect — recasting transparency as a communicative right rather than a technical or compliance property. Implications for research, platform governance, and inclusive design are discussed.