Poster
in
Affinity Event: Queer in AI
Depictions of Queer Mental Health by Grok-2
Declan Grabb
Keywords: [ lgbt; mental health; image generation; multi-modal ]
Mental health disorders have long been stigmatized by society, with historical attributions to supernatural causes and, more recently, persistent biases against marginalized groups. This stigmatization continues to hinder care-seeking behavior and negatively affects health outcomes, particularly for LGBTQ+ individuals whose identities have been pathologized in diagnostic criteria. Media representations of mental illness, such as the controversial portrayal of suicide in "13 Reasons Why," have been linked to real-world harm, emphasizing the influence of imagery on public perception. With the rise of image-generating AI models, the portrayal of mental health symptoms may either perpetuate existing stereotypes or serve as a tool to reduce stigma. This study explores how the AI model Grok-2 visualizes various mental health symptoms both in isolation and when contextualized with specific demographic information, revealing potential biases in depiction. Through this examination, I aim to understand how AI's portrayal of mental health might impact public perception and stigma, advocating for thoughtful and respectful use of AI in mental health discourse.
Live content is unavailable. Log in and register to view live content