Responsible Clinical AI in Dentistry: Trust, Professional Autonomy, and Perceptions of Accountability Across Stakeholders in Romania—A Multidisciplinary Cross-Sectional Survey

Background/Objectives: Artificial intelligence (AI) is entering clinical dentistry more quickly than the professional, ethical, and regulatory frameworks intended to govern it. Responsible implementation depends not only on diagnostic performance but on trust, professional autonomy, and clear accountability—dimensions rarely examined across the different stakeholder groups who shape dental care. This study assessed perceptions of trustworthy AI, accountability, and regulatory preparedness among dental and non-dental stakeholders in Romania. Methods: A multidisciplinary cross-sectional survey using purposive, non-probability sampling, reported in accordance with the CROSS guideline, was administered to dental professionals, legal professionals, forensic physicians, and other healthcare stakeholders. The instrument addressed AI literacy, governance and accountability, ethics and transparency, and clinical implementation. Responses were compared across professional groups using chi-square and Kruskal–Wallis tests with Benjamini–Hochberg or Holm correction, and the questionnaire structure was examined through an exploratory psychometric assessment. Of 304 submissions, 287 were retained for analysis. Results: Agreement on ethical and governance safeguards was high (84–91%), including testing on diverse populations, independent auditing, and patient disclosure. Between-group differences concentrated in AI awareness, literacy, and educational exposure, as well as in governance and accountability, whereas ethics and transparency items showed broad consensus. Respondents attributed responsibility for AI-related outcomes to clinicians (77.4%), developers (49.5%), and healthcare institutions (28.3%). Only a minority reported AI Act awareness (32.8%) or previous participation in AI-related courses or conferences (25.2%). Conclusions: Romanian stakeholders included in this multidisciplinary sample share strong normative expectations for trustworthy clinical AI but report limited regulatory awareness and educational exposure. These findings support the development of structured AI preparedness initiatives and clearer governance frameworks reflecting stakeholders’ perceptions of responsibility as extending across clinicians, developers, and healthcare institutions. Professional certification before clinical AI use emerged as a specific stakeholder preference.