Disinformation in the digital era: The role of deepfakes, artificial intelligence, and open-source intelligence in shaping public trust and policy responses

Abstract

This study investigates the role of deepfake and open-source intelligence (OSINT) in enabling disinformation campaigns and their societal consequences. Using the Deepfake Detection Challenge (DFDC) dataset for technical evaluation, social media datasets for OSINT network and sentiment analysis, and public opinion data from the Global Disinformation Index, the study applied machine learning classification, network analysis, sentiment analysis, and interrupted time series (ITS) analysis. The technical assessment achieved a detection accuracy of 0.73, precision of 0.75, and recall of 0.70, identifying areas for enhancement in identifying synthetic media. OSINT analysis revealed pivotal amplifiers of disinformation, with User1 having a degree centrality of 0.263 and betweensess centrality of 0.135. Sentiment analysis showed an average sentiment score of -0.085, while ITS analysis documented a significant 9.76-point decline in public trust post-disinformation events. Recommendations include developing adaptive AI detection systems, implementing global regulatory measures, fostering public media literacy, and encouraging ethical OSINT practices. Keywords: Deepfakes, Artificial Intelligence, Disinformation Campaigns, Open-Source Intelligence, Public Trust.