<![CDATA[Background: Generative AI has intensified the circulation of synthetic visual content and increased the need for users to identify AI-generated content. Prior studies have examined deepfake detection, AI literacy, AI experience, and social interaction separately, but empirical evidence on how these factors jointly shape identification ability remains limited. Objective: This study investigates the effects of AI Social Interaction and AI Experience on AI Content Identification Ability, with AI Literacy positioned as a mediating mechanism. Methods: A quantitative survey was conducted with 270 social media users in Jabodetabek. AI Social Interaction, AI Experience, and AI Literacy were measured using a five-point Likert questionnaire, while identification ability was measured through a 30-item objective performance test consisting of AI-generated and authentic images and videos. Data were analyzed using partial least squares structural equation modeling with SmartPLS 4.0. Results: AI Experience had a positive effect on AI Literacy. AI Literacy and AI Social Interaction had significant direct effects on AI Content Identification Ability. AI Experience did not directly affect identification ability, but its indirect effect through AI Literacy was significant, indicating full mediation. Conclusion: AI use experience does not automatically improve the ability to identify synthetic content. Experience must be converted into evaluative and ethical AI Literacy, while social interaction may support direct pattern recognition through digital exposure.]]>
