Generative AI tools have become increasingly accessible to the general public during the last few years, empowering users to create images with unprecedented ease. However, these tools are often trained on biased datasets, which can lead to the production of biased imagery, an underlying issue that often goes unnoticed by users. This problem is particularly concerning in the design field, where advertising agencies and other companies are progressively integrating AI-generated visuals into their workflows, resulting in inadvertently reinforcing stereotypes in visual outputs. We conducted a study involving 114 participants, to uncover how societal biases manifest when people are presented with AI-generated images of individuals from various ethnic and cultural backgrounds, seeking to uncover patterns of racial profiling and the underlying mechanisms of crime racialization, and how this can harm the creation and selection of AI-generated images in the design industry. The findings highlight a significant and pervasive issue, raising important questions about the ethics and implications of generative AI, and the potential harm this could cause to the design field. This research is of particular interest to educators, designers, AI developers, and anyone who uses or encounters AI-generated content.