TRUST IN ARTIFICIAL INTELLIGENCE-GENERATED HEALTH INFORMATION AND RECOMMENDATIONS: A CROSS-SECTIONAL SURVEY OF 514 OVERWEIGHT AND OBESE OUTPATIENTS

Furkan TÜRKOĞLU, Elif Nur GENCER, Emre ERDOĞAN, Soner SARI, Alperen YİĞİT, Özgür DOĞAN

Anatolian Current Medical Journal - 2026;8(5):922-928

Department of General Surgery, Aktif International Hospital, Kocaeli, Turkiye

 

Aims: To evaluate the perceptions of overweight and obese patients attending an outpatient clinic regarding the use of Artificial Intelligence (AI) for health-related issues, their trust in AI-generated information on obesity-related musculoskeletal complaints, and their attitudes toward receiving exercise and nutrition recommendations from AI. Methods: This cross-sectional, anonymous survey included 514 overweight and obese patients attending outpatient clinics. The questionnaire consisted of demographic questions and six items related to AI. Associations between demographic characteristics and survey responses were analyzed. Results: Overall, 49.4% of participants reported having used AI for health-related purposes at least occasionally. The most common response regarding trust in AI-generated health information was "I'm not sure" (59.3%). Participants were significantly more willing to receive nutrition recommendations than exercise recommendations from AI. Nearly half of the participants (49.4%) believed that physicians would continue to play an essential role in healthcare despite advances in AI. AI use, perceived usefulness of AI, and trust in AI were significantly associated with both age and educational level (all p<0.01). By gender, only willingness to receive AI-generated exercise recommendations differed significantly (p=0.017). Conclusions: Patients were generally inclined to accept AI as a supportive tool under physician supervision, particularly for nutritional guidance and health information. However, they expressed greater reservations about AI-generated exercise recommendations and demonstrated a clear preference for professional validation. Age and educational level were significantly associated with these attitudes.