Mitigating mental illness stigma through narrative identification: The impact of AI authorship, emotionalization, and perspective

This experimental study investigates whether Artificial Intelligence (AI) authored narratives could mitigate the stigmatization of people with depression. Three messages written about the hardship of people living with depression were manipulated in terms of emotionalization (high vs. low), perspective (first vs. third person), and authorship label (AI, human, none). The dependent variables measured were identification and destigmatization. While only a few expected main and interaction effects emerged, a moderated mediation model revealed a more intricate set of relationships among the experimental variables. These findings highlight the nuanced conditions under which AI-generated content can meaningfully influence attitudes, suggesting its promising potential as a tool for stigma reduction.