Triage processes are a crucial function of healthcare providers' practice, such as physicians and nurses, that establishes the foundation for ensuring appropriate and timely care for all patients in healthcare settings, prioritizing and addressing the severity of the presenting problem (Charan, Kalia, Dular, Kumar, & Kaur, 2025; Da'Costa et al., 2025). Efficient and accurate triage decision-making process is essential to optimize patient outcomes and reduce wait time in the clinical settings (Agius, Magri, & Cassar, 2025; Kassa et al., 2025). The increase in the number of people seeking care in the various clinical settings and constraints in healthcare systems has put substantial pressure on triage processes (Al-Ghabeesh, Thabet, Rayan, & Abu-Snieneh, 2023; Almulihi et al., 2024; Alnaeem, Islaih, Abu Sabra, & Bani-Hani, 2025). Traditionally, triage has depended on human skills in judgment, which are vulnerable to errors and limitations, leading to potential misclassification of patient acuity. (Gellert et al., 2024; Liu et al., 2025).

Artificial intelligence (AI) is a transformative tool that has potential for great power in the healthcare system, offering potential improvements in accuracy, efficiency, and consistency of triage decisions, including diagnosis, treatment, and the clinical decision-making process (Arnaud, Elbattah, Moreno-Sánchez, Dequen, & Ghazali, 2023; Aryffin, Noor, Musa, & Baharuddin, 2024; Kuttan, Pundkar, Gadkari, Patel, & Kumar, 2025). In this context, incorporating AI technologies has emerged as a promising aid in assisting and supporting healthcare providers in the triage decision-making process at clinical settings by providing a more accurate, efficient, and consistent assessment process, analyzing patient data, and predicting severity levels that could improve the outcomes (Araouchi & Adda, 2025; El Arab & Al Moosa, 2025; Preiksaitis et al., 2024).

Several scientists have developed AI models to assist in the triage decision-making process. For instance, the development of i-TRIAGE by Kipourgos, Tzenalis, Koutsojannis, and Hatzilygeroudis (2022) is consistent with the growing body of research exploring AI applications in triage in emergency healthcare settings. By demonstrating the effectiveness of AI techniques, such as machine learning (ML), in enhancing triage performance, i-TRIAGE—which is supported by the widely recognized Emergency Severity Index (ESI) protocol—is intended to assist triage nurses in making prompt and accurate decisions regarding patient urgency and appropriate specialist referral. Likewise, the ChatGPT-based triage system demonstrated promise in high-acuity cases with a 76.6 % accuracy rate and good agreement for ESI levels 1 and 2 (Colakca et al., 2024).

Another recent advance in AI is the development of virtual triage (VT) systems, capable of assessing patient symptoms, recommending care levels, and improving healthcare navigation. Gellert et al. (2024) made a substantial contribution to this growing field by evaluating a mixed model: AI-driven VT integrated with a live nurse triage service. This model tackles enduring problems in healthcare, including underutilization of lower-acuity services, wasteful ED visits, and mismatched care acuity. Utilizing a sizable dataset (N = 54,587) spanning 26 months gives the study more statistical power and practical relevance. It builds upon prior research suggesting that AI can enhance the patient decision-making process and supports findings from similar interventions where VT systems have led to better resource allocation and more appropriate care-seeking behavior.

Despite the increasing prevalence of AI, researchers are still exploring its potential applications for physicians and nurses. For instance, a recent review was conducted by Sengar, Hasan, Kumar, and Carroll (2025), who found that AI-powered triage systems present an incredible opportunity to enhance patient outcomes and healthcare delivery. Another recent systematic review found that over 60 % of medical experts have stated that they are hesitant to use AI systems because they are afraid of data insecurity and lack transparency (Khan, Shah, Shaikh, Thabet, & Belkhair, 2025). Therefore, this review's novelty focused on how healthcare providers can use it in various settings, time frames, or areas that require attention, as well as evaluation and understanding of AI's role, limitations, challenges, and ethical concerns. This will be a continuous investigation, with questions about its effectiveness, dependability, and real-world applicability in clinical practice. The purpose of this systematic review was to synthesize the studies that explore AI's role in enhancing healthcare providers' triage decisions process in the various healthcare setting. The specific aims of the study were to: (1) evaluate the benefits, accuracy, and efficiency of AI-assessed triage compared to traditional methods; (2) identify the challenges, ethical considerations, and limitations associated with AI triage system implementation; and (3) provide a comprehensive understanding of how AI can be leveraged to improve triage practices and inform future research and policy in this evolving field.