Breaking the illusion: impact of news literacy on deepfake identification and sharing

Purpose: Deepfakes pose a serious threat to societies. While deepfake-specific literacy programs are emerging, they may backfire and reduce trust in media. This study aims to explore traditional news literacy as an alternative buffering factor. We examine how news literacy can combat the effects of deepfakes and whether deepfake knowledge and heterogeneous political discussion moderate this effect. Design/methodology/approach: We conducted a cross-sectional survey with a representative sample in the United States (N = 1,008) using Qualtrics. We run bootstrapped regression mediation and moderation analysis to test our hypotheses. Findings: Results suggest that higher news literacy is associated with lower deepfake sharing by helping individuals discern deepfakes. Those with greater deepfake knowledge and more frequent heterogeneous political discussions are better at using news literacy to identify deepfakes and avoid sharing them. Originality/value: This study is among the first to examine how traditional news literacy can counter deepfakes, a growing form of disinformation. It extends news literacy theory by assessing its role in protecting against multimodal disinformation. Additionally, it contributes to cognitive theory by highlighting the combined influence of news literacy, deepfake knowledge, and social media discussion heterogeneity in resisting deepfakes. These insights offer practical strategies to reduce the harm caused by deepfakes.