Assessing the “Optimism–Knowledge Gap”: An Exploratory Study of AI Awareness, Application, and Educational Needs Among a Sample of Italian Clinicians

Background: Artificial intelligence (AI) is poised to fundamentally reshape healthcare delivery, offering unprecedented advancements in diagnostics, treatment personalization, and operational efficiency. However, a growing body of international research reveals a critical “optimism–knowledge gap”: healthcare professionals are enthusiastic about AI’s potential but possess limited technical knowledge and practical experience. This gap compromises the safe and effective implementation of AI tools. The Italian healthcare context presents a unique and amplifying challenge, as it is defined by the stringent “human-in-the-loop” oversight mandated by the Garante per la protezione dei dati personali (Italy’s Data Protection Authority). This legal framework makes clinician competence not just a goal, but a prerequisite for regulatory compliance. Objective: This study aimed to provide an exploratory quantitative assessment of AI awareness, practical application, and understanding of its limitations among a sample of clinicians in Italy. It specifically sought to compare the preparedness of hospital-based clinicians and general practitioners (GPs) and to identify the workforce’s perceived educational needs within this unique legal environment. Methods: A descriptive, cross-sectional survey was conducted from February to August 2025. Using a non-probability convenience sampling method via professional networks, the survey yielded 362 total responses. Data were analyzed descriptively and inferentially using Chi-square (χ2) tests to compare cohort responses on familiarity, practical exposure, knowledge of limitations, and interest in further training. Results: A universal and high demand for education was found, with 89.9% of all respondents being “Moderately” or “Very” interested in learning more about AI. This optimism coexists with dangerously low practical exposure. The gap was most profound among GPs, 44.1% of whom have “Never” used an AI tool—a rate significantly higher than hospital clinicians (34.9%; χ2=3.14, p = 0.045). Furthermore, 32.6% of GPs admitted that they “understand some benefits but not the limitations.” Conclusions: Italian clinicians mirror the global optimism–knowledge gap. These findings underscore the urgent need for structured, continuous education in AI literacy to address ethical and regulatory imperatives within the Italian healthcare system.