Simulation-based analysis of AI responses to sexual myths during pregnancy
Year: 2026
Type: article
Abstract: Aims: This study aimed to evaluate the quality, scientific accuracy, and communication characteristics of artificial intelligence generated responses to common sexual myths during pregnancy using a simulation-based approach. Methods: Ten simulated patient questions reflecting widespread, non–evidence-based beliefs about sexual activity during pregnancy were deliberately constructed based on themes frequently reported in the literature. These questions addressed fetal perception, physical harm, developmental consequences, and moral or religious concerns. Responses generated by ChatGPT were analyzed qualitatively according to predefined criteria, including scientific accuracy, internal consistency, communication tone, and potential clinical risk. No real patient data were used, and no human subjects were involved. Results: All responses were found to be broadly consistent with established obstetric and fetal physiology principles. The system consistently provided clear, reassuring, and nonjudgmental explanations, particularly for concerns related to fetal safety. However, individualized obstetric risk factors were rarely emphasized, which may limit clinical applicability in certain high-risk contexts. Responses to morally or culturally sensitive questions demonstrated a cautious and neutral communication style. Conclusion: Artificial intelligence generated responses may serve as a supportive informational tool for addressing common sexual myths during pregnancy. While scientifically accurate and communicative, such systems should not replace individualized clinical counseling. Their greatest value may lie in complementing physician–patient communication by identifying prevalent misconceptions and facilitating early educational interventions.
DOI: 10.51271/jcogp-0064
Author: Seniye Burcu Torumtay Alıç
Language: English
Subfield: Health Informatics
Field: Medicine
Domain: Health Sciences
Open Access status: hybrid
OpenAlex ID: W7127064971