From Truth To Synthetic Reality: Deep Fakes, Media Ethics And The Erosion Of Digital Trust In The Age Of Artificial Intelligence

Abstract

The rapid diffusion of generative artificial intelligence has made synthetic media — commonly known as deepfakes — cheap to produce, difficult to detect, and increasingly consequential for public life. Drawing on 2024 industry surveys, academic meta-analyses, and cross-national trust barometers, this thesis examines how synthetic media is reshaping media ethics and eroding digital trust. It synthesises data from Regula/Sapio Research, Sumsub, Entrust, Deloitte, the Reuters Institute Digital News Report, Gallup, and a 2024 peer-reviewed meta-analysis of human deepfake-detection performance. The findings show that nearly half of surveyed businesses experienced deepfake-related fraud in 2024, that election-related synthetic media surged by triple- and quadruple-digit percentages across dozens of countries, and that ordinary people detect high-quality deepfakes only marginally better than chance, while public trust in news held at historic lows. The thesis argues that this convergence of technical capability, weak detection, and declining institutional trust constitutes a structural crisis for journalism, law, and democratic deliberation, and it closes with a framework of ethical, regulatory, and technical responses grounded in the evidence reviewed.