An Immediate, All-Knowing Provider: How Artificial Intelligence Responds to a Pregnant Patient Regarding Acetaminophen Use [ID 3689]

INTRODUCTION: As technology progresses, the utility of artificial intelligence (AI) may connect patients to medical information. Recent debate has cast doubt on the safety of medications during pregnancy, with acetaminophen at the forefront. This study sought to determine the behavior of an AI software when questioned by a presumed pregnant individual. METHODS: Five queries were submitted to a chatbot (ChatGPT-5), and answers were correlated by alignment with primary, science-based societies (ie, the American College of Obstetricians & Gynecologists and the Society for Maternal-Fetal Medicine). Prompts included relationship between acetaminophen and autism, safety of dosage, common ailments for use, research studies regarding the dangers of intake during pregnancy, and highlighting individuals on social media. Responses were compared to statements by the aforementioned societies. RESULTS: ChatGPT offered discussion ranging from the absence of to possible associations between acetaminophen and autism; however, a final statement insisted “no studies prove acetaminophen is harmful when used in normal doses during pregnancy.” The chatbot listed common ailments for drug use, indicated the importance of appropriate treatment, and stressed the use of the lowest effective dose for the shortest time (maximum of 3,000 mg daily). ChatGPT demonstrated the lack of proof between acetaminophen and pregnancy outcomes by analyzing and referring to scientific studies while spotlighting board-certified, science-practicing obstetrician–gynecologists. CONCLUSIONS/IMPLICATIONS: Despite conflicting arguments regarding acetaminophen safety in pregnancy, for general questions, AI may be utilized to steer patients toward science-based research. Answers were medically accurate, simple, and referenced trusted organizations. This also highlights the importance of medical provider involvement in social media, as even language models may harvest and provide resources to the public.