Explore how generative AI systems inadvertently suppress identity-related speech through their content moderation mechanisms in this 26-minute conference presentation from ACM's best papers session. Examine the research findings by Grace Proebsting, Oghenefejiro Anigboro, Charlie M. Crawford, Danaé Metaxa, and Sorelle A. Friedler that reveal how automated content moderation systems may disproportionately filter or restrict speech related to specific identities. Discover the implications of these suppression patterns for marginalized communities and understand the technical and ethical challenges that arise when AI systems make decisions about identity-related content. Learn about the methodologies used to identify these biases in generative AI platforms and consider the broader consequences for free expression and equitable access to AI-generated content creation tools.