Epistemic Trust in AI-Mediated Social Media: An Integrative Review of Young Adults’ Encounters with Misinformation

Abstract

Generative artificial intelligence (AI) is changing how young adults encounter, produce, and assess information online. This integrative review examines how AI-mediated communication intersects with epistemic trust and misinformation, with particular attention to emerging adults who use social platforms for information and social interaction. Rather than presenting new empirical data, the article synthesizes recent peer-reviewed research and selected cross-national evidence published mainly from 2020 to 2026. The literature points to a dual problem. Generative systems can produce fluent but inaccurate material and can lower the cost of creating persuasive synthetic content; at the same time, AI-assisted explanation, correction, and detection may help users evaluate claims. Trust is therefore not simply increased or decreased by AI. It is shaped by source cues, disclosure, perceived expertise, prior beliefs, platform design, and the availability of independent verification. Young adults’ familiarity with digital media should not be equated with reliable misinformation detection, and blanket distrust can be as damaging as uncritical acceptance. The review develops a framework connecting system characteristics, social context, individual judgment, and institutional accountability. It argues that responses should combine specific provenance and verification cues, media and AI literacy, accessible independent sources, and transparent platform governance. The evidence base remains uneven across countries, languages, and age groups; accordingly, the article identifies testable questions for comparative research rather than claiming universal effects.