TikTok has become a central arena in which Generation Z users construct and negotiate digital identity, yet research on how its affordances relate to this process remains fragmented across platform-feature, algorithmic, and self-presentation perspectives. This article presents a systematic literature review that examines how TikTok affordances have been conceptualized, how existing studies explain their role in Generation Z digital identity construction, and what theoretical, methodological, and empirical gaps remain. Following PRISMA 2020 guidance, 37 peer-reviewed studies published between 2021 and 2026 were retrieved from Scopus and a supplementary Google Scholar search, appraised using the Mixed Methods Appraisal Tool, and analyzed through thematic synthesis, yielding five descriptive and three analytical themes. The review finds that algorithmic affordances (recommendation, personalization, and visibility) dominate the literature, while social affordances remain comparatively under-examined. It identifies three mechanisms linking affordances to identity: algorithmic recognition, the negotiation of visibility and platform power, and community-based affirmation and cultural meaning-making. On this basis, the review reconceptualizes TikTok as an identity infrastructure rather than merely a platform for self-presentation. This reconceptualization is offered as a conceptual proposition rather than a demonstrated effect. Because 33 of the 37 studies rely primarily on qualitative and self-reported accounts of user experience, and relatively few directly examine recommendation systems or their effects over time, the evidence is better suited to explaining how users perceive and interpret algorithmic influence than how recommendation systems objectively shape identity. The literature is also concentrated on marginalized communities. Advancing the field, therefore, requires observational, computational, and longitudinal research extending beyond the populations studied to date.