Healthcare Professionals’ Perceptions of AI-Assisted Clinical Decision-Making in Jordan: A Qualitative Study of Trust, Accountability, System Readiness, and Professional Practice

Highlights What are the main findings? Healthcare professionals in Jordan expressed cautious but constructive views toward AI-assisted clinical decision-making. Trust in AI was conditional and depended on local validation, transparency, institutional endorsement, and clear accountability. Participants supported AI as an adjunct to clinical judgement, particularly for screening, risk flagging, and information support. Medication-related safety emerged as a cross-cutting concern, particularly where AI recommendations could influence dosing, drug interactions, polypharmacy, or therapeutic monitoring. Participants linked safe AI implementation to governance, ethical data practices, workforce training, local validation, and equitable digital infrastructure. What are the implications of the main findings? AI implementation in Jordanian healthcare should be approached as supervised clinical decision support rather than autonomous decision-making. Participants’ accounts suggest the need for local validation, clear accountability procedures, ethical data governance, and institutional policies defining how AI outputs are used, documented, challenged, or overridden. Interprofessional AI training is needed to help healthcare professionals critically appraise AI outputs, recognise limitations, and avoid over-reliance. Where AI tools influence medication-related decisions, medication-safety expertise, including pharmacist input where appropriate, should be considered. Abstract Background/Objectives: Artificial intelligence (AI) is increasingly used in clinical decision-support systems, yet its adoption in low- and middle-income countries, including Jordan, remains limited and underexplored. Understanding how healthcare professionals perceive AI-assisted clinical decision-making is essential for safe and contextually appropriate implementation. This study explored healthcare professionals’ perceptions of AI-assisted clinical decision-making in Jordan, with particular attention to trust, accuracy, accountability, professional judgement, digital literacy, and health-system readiness. Medication-related safety and prescribing concerns were examined as secondary cross-cutting issues where they emerged from participants’ accounts. Methods: A qualitative study was conducted using semi-structured, in-depth interviews with 22 purposively sampled healthcare professionals from public, private, and university-affiliated healthcare institutions in Amman, Irbid, and Zarqa. Participants included physicians, nurses, pharmacists, and allied health professionals with varied specialties and levels of seniority. Data were analysed using Braun and Clarke’s reflexive thematic analysis. Member checking, peer debriefing, reflexive memos, and audit trails were used to enhance trustworthiness, and reporting followed the Consolidated Criteria for Reporting Qualitative Research (COREQ). Results: Eight overarching themes were identified: conditional trust in AI-assisted clinical decision-making; concerns regarding accuracy and confident algorithmic errors; accountability and professional responsibility; AI as an adjunct rather than a substitute for clinical judgement; the influence of experience, specialty, and digital literacy on AI acceptance; Jordanian health-system readiness; privacy, confidentiality, and algorithmic bias; and training requirements for safe AI use. Medication-related safety emerged as a cross-cutting concern, particularly in relation to dosing, polypharmacy, drug–drug and drug–herb interactions, and the risk of over-reliance on AI-generated recommendations. Conclusions: Healthcare professionals in Jordan expressed cautious but constructive views toward AI-assisted clinical decision-making. AI was perceived as potentially useful when used to support, rather than replace, professional judgement. Participants’ accounts suggest that safe implementation depends on local validation, clear accountability frameworks, ethical data governance, interprofessional training, and careful consideration of medication-safety expertise where AI tools influence prescribing or therapeutic decisions. These findings highlight the importance of context-sensitive AI governance strategies that support trustworthy, accountable, and professionally supervised AI adoption in healthcare.