Optimistic and ambivalent attitudes toward clinical artificial intelligence

Artificial intelligence (AI) is increasingly being integrated into healthcare systems; however, healthcare professionals' knowledge, attitudes, and perceptions regarding AI remain insufficiently understood. This study aimed to evaluate healthcare professionals' perceptions of AI in clinical practice.

This descriptive cross-sectional study included 260 healthcare professionals, including physicians, nurses, and healthcare technicians working at a tertiary public hospital in Türkiye. Participants completed the Healthcare Professionals and Artificial Intelligence Questionnaire, which assessed demographic characteristics, AI knowledge, workplace AI use, attitudes toward AI applications, ethical and legal concerns, perceived professional impact, and educational needs.

Most participants reported limited AI knowledge; 64.6% indicated basic knowledge, whereas 23.1% reported no knowledge. Only 24.6% reported using AI-based systems in clinical practice, while 35.4% were uncertain whether such systems were used in their workplaces. Despite limited experience, attitudes toward recent AI developments were generally positive. Most participants expressed willingness to receive AI-related education (76.9%), although only 6.2% had previously received formal training. Nearly half of the participants believed that the human factor would remain essential in healthcare, whereas a similar proportion considered partial replacement by AI possible. Physicians demonstrated greater confidence in AI usefulness and stronger willingness to integrate AI into clinical decision-making than nurses and other healthcare professionals. Nurses primarily emphasized data security and privacy concerns, whereas physicians highlighted patient safety and misdiagnosis risk. Exploratory psychometric analyses demonstrated a three-factor structure with acceptable sampling adequacy (KMO = 0.765).

Healthcare professionals generally perceive AI as a supportive tool rather than a replacement for human judgment. Limited AI literacy, uncertainty regarding legal responsibility, and insufficient training remain important barriers to responsible AI adoption. Profession-sensitive educational strategies and clearer governance frameworks may support safer AI integration into healthcare practice.