Summary: The rapid integration of generative AI tools into medical publishing has introduced new opportunities and risks, particularly in the creation of visual materials, such as anatomical illustrations. This study assessed the prevalence and characteristics of AI-generated facial anatomy images across peer-reviewed journals and online course promotional materials in plastic surgery and cosmetic dermatology. A total of 4734 manuscripts were screened, yielding 37 articles containing facial anatomical depictions. AI-generated content was identified in 10.8% (4 of 37) of relevant articles published in 2025, representing a marked increase from 2024, with all identified images demonstrating gross anatomical inaccuracies. A complementary web search of online anatomy courses also found that nearly half of the sites used erroneous AI-generated imagery in promotional content. These findings underscore growing concerns over the rapid and unregulated adoption of powerful AI tools for visual content in medical publishing, with regard to scientific misinformation, legal risks, and professional standards. In this special topic discussion, we describe how generative AI models function, outline their limitations, and explore the ethical, legal, and professional implications of their misuse. To ensure the safe and responsible integration of visual AI generators in medical publishing, the urgent adoption of clear standards and oversight protocols should be considered for anatomical and surgical depictions.
Zhang et al. (Wed,) studied this question.