<![CDATA[The rapid improvement of generative artificial intelligence has made deepfake images, audio, and videos increasingly realistic and difficult for ordinary users to evaluate. This study examines the effectiveness of media literacy in improving the public’s ability to detect deepfake content. A qualitative systematic literature review was conducted using a structured Deep Search workflow covering publications from 2021 to 2026. the broad retrieval stage identified 8,190,643 records related to the topic, followed by targeted searches on intervention design, critical thinking, digital literacy, artificial intelligence literacy, and null or harmful effects. Relevance screening produced 240 candidate records, 144 eligible records after deduplication and thresholding, and 50 studies for qualitative synthesis. the findings show a mixed but generally positive evidence base. Media literacy is most effective when it emphasizes analytical reasoning, concrete examples, repeated practice, immediate feedback, and verification habits. Generic definitions, brief warnings, and passive exposure frequently produce weak or null effects. Some interventions also create a trust trade-off by improving fake-content detection while increasing suspicion toward authentic media. Individual factors, including critical thinking, self-efficacy, prior exposure, and cognitive bias, interact with content quality and platform context. the review concludes that targeted and interactive media literacy can improve deepfake discernment, but it should be combined with confidence calibration, multimodal verification, and technological support to preserve trust in authentic information.]]>
Copyrights © 2026
