Misidentification in a Digital Age: The Ethical Implications and Challenges of AI Integration in Scientific Illustration

Scientific illustration is a crucial resource in scientific communication, providing visual educational assistance for fields such as science, medicine, and agriculture. This thesis explores the influence of artificial intelligence (AI) on scientific illustration and examines the strengths and weaknesses of AI-generated scientific illustration as compared to human-made illustration. Through the analysis of peer-reviewed sources, this study brings awareness to the common issues presented in AI-generated scientific images, explaining the risks associated with professional, research, and public settings. To address these challenges, research is presented to explain tactics to mediate potential errors associated with AI-generated images, the importance of human intervention and observational skills, and present a visual solution to further compare human-made and AI-generated scientific illustrations. By juxtaposing AI-generated illustrations with human-made examples, the visual solution and research will further bring awareness to professionals, researchers, and the public alike, with the goal to reduce the risk of AI-related harm and misdiagnosis in scientific and medical fields of study. This research aims to bring an understanding of the potential for AI-generated illustrations in scientific fields without negating the importance and relevance of human-made scientific illustration.