When Health Messages from Someone Look Like You: Deepfake-Enabled Similarity, Perspective-Taking, and the Moderating Role of AI Literacy

Abstract

This paper examines self-relevance cues in narrative-based health-promoting messages by integrating deepfake-enabled facial similarity as a technology-based, perceptual-level cue and perspective-taking as an imaginative, message-level cue. Results from a lab experiment (N = 204) showed that both high facial similarity via deepfake technology and first-person perspective significantly increased self-referencing and persuasion. Identification was influenced solely by first-person perspective-taking, with no direct effect of deepfake similarity. Importantly, three-way interactions with AI literacy emerged for identification and persuasion: among high-AI literacy audiences, first-person framing compensated for weak perceptual similarity, whereas its influence diminished when similarity was high. In contrast, among audiences with low AI literacy, the strongest effects occurred when high similarity was paired with first-person framing. These results broaden our understanding of personalized health communication by demonstrating how AI-driven technological and message-level cues jointly—but contingently—shape self-relevance processes in AI-mediated health communication.

Ähnliche Arbeiten