"Should I trust or Should I not?" A Survey to measure AI Confidence Levels Among Students of the bachelor's degree in Psychiatric Rehabilitation Techniques and Psychiatry Residents. Introduction

The use of Artificial Intelligence (AI) in mental health context has significantly expanded in recent years, offering innovative solutions in various clinical settings. This study investigates the levels of trust and familiarity with AI among psychiatric rehabilitation students (PRTS) and psychiatric residents (PR) at the University of Modena and Reggio Emilia. Methods: An online questionnaire developed by the University of Twente was administered using the REDCap platform to collect data from 78 participants, including 53 PRTS and 25 PRs. Results: The findings revealed that 80% of PRTS reported familiarity with AI, compared to less than 50% of PRs. However, only 42.6% of PRTS and 22.5% of PRs felt familiar with AI chatbots. Trust in AI-driven recommendations was generally neutral across both groups, with 40.8% of respondents expressing neither agreement nor disagreement. Additionally, over 68% of PRTS and 70.8% of PRs expressed confidence in their ability to generate high-quality ideas. Conclusion: A more structured AI education within medical training is needed to bridge the gap between familiarity and trust. AI, with its continuous advancements, offers substantial potential in healthcare, bringing with it both opportunities and risks.