Deepfakes in the Media Ecosystem: Synthetic Information and Its Impact on Digital Audiences

Abstract

This article examines the deepfake phenomenon, not as an isolated technological
innovation, but as a structural element of the modern media landscape, integrated into the logic
of digital production and emotional audience engagement. The aim of the study is to analyze the
impact of AI-generated media content on information perception and the formation of trust in
media, as well as to develop a typological approach to analyzing deepfakes. The methodological
basis of the work is the concepts of media trust, digital disinformation, and synthetic media, as
well as elements of empirical analysis used to classify various forms of deepfake content. The
study demonstrates that deepfakes represent a systemic problem in the media environment,
increasing the risks of manipulation, propaganda, and disinformation, and contributing to
the erosion of digital audience trust in media content. The need to classify deepfakes not only
by technical characteristics, but primarily by their degree of impact and semantic purpose,
including manipulative, propaganda, disinformation, and ironic forms, is substantiated. The
study concludes that existing journalistic and editorial practices for countering synthetic media,
including verification tools, are limited in effectiveness without changing audience attitudes
toward the information they consume and disseminate. The sustainability of the information
environment in a digital society requires a redistribution of responsibility between media
professionals and users, as well as the development of preventive journalism and analytical
formats.