Generative AI systems like Stable Diffusion and DALL·E have transformed creative and computational fields by enabling the automated production of synthetic images, text, and multimedia content. With millions of people relying on AI daily for work, education, and personal use, these technologies are increasingly shaping digital representation and media. Despite these advancements, concerns about bias and harmful portrayals persist. In this study, I examine the social and representational factors that contribute to inaccurate and dehumanizing portrayals of Black women in AI-generated images using qualitative methods such as comparative visual analysis and thematic coding. To investigate this, I generated AI images depicting White and Black women engaged in everyday activities. To do this, I used a Stable Diffusion codebase to create images based on text input; I then analyzed these images through racial and gender comparisons to assess inaccuracies and biases. Preliminary results reveal discrepancies in the depiction of Black women compared to White women performing the same tasks. Images of Black women were more frequently oversexualized, depicted with less context-appropriate clothing and exaggerated body features. These patterns suggest underlying biases in the training data. To further explore these disparities, I will generate images across diverse scenarios and apply qualitative coding and training data evaluation for deeper analysis. This study underscores the need for more equitable AI training data and stricter bias mitigation strategies, contributing to responsible AI governance and policy development.

Over the years, progress has been made in the world of education. New strategies have been developed to reach students' potential. However, despite all of the progress, there is still a gap between low-income students' academic success rates versus their more affluent peers. While there are many ideas for why we are seeing this gap, my research focuses on informing educators on how to support low-income students. The research aims to understand the academic success gap between low-income students and their more affluent peers and to determine supports that may offset this gap. The particular focus of the research aims to gain a deeper understanding of why we are seeing a gap in academic success rates. This research seeks to find ways to inform educators on how to support low-income students. The study will focus on identifying the support low-income students need to improve their academic success. To identify these supports, I will seek the opinions of Wisconsin middle school teachers who work with low-income students. In this study, I will be reviewing the research on the academic success seen in low-income students in comparison with their more affluent peers. Through my literature review, I will identify the barriers faced by low-income students across the United States along with resources to offset the barriers. With this information, I will conduct interviews with middle school educators in Wisconsin. Through this investigation, I hope to find stronger resources that will aid low-income, middle school students in Wisconsin in achieving better academic success.