Artificial intelligence (AI) is increasingly recognized as a transformative force in healthcare, with growing relevance in palliative care. This article examines the clinical potential, current applications, risks, and ethical preconditions associated with AI implementation in this sensitive field. AI-driven systems enhance personalized symptom management by analyzing large datasets derived from electronic health records (EHRs), patient-reported outcomes, and clinical assessments. Machine learning algorithms identify patterns in symptom trajectories and treatment responses, enabling individualized care plans. Reported median prognostic accuracies range between 78% and 83% for survival prediction in advanced illness populations, while prediction of treatment response and pain management outcomes achieves approximately 80–85% accuracy. AI applications also contribute to caregiver support through chatbots and digital platforms providing continuous informational and emotional assistance, and to system-level improvements via symptom-tracking applications, virtual reality tools, and AI-supported care coordination systems. Furthermore, AI strengthens research capacity by enabling large-scale data analysis and identifying novel risk factors, such as delirium prediction models with sensitivity up to 75% and specificity up to 88%. Despite these advantages, implementation raises ethical and practical concerns, including data privacy risks, algorithmic bias, model inaccuracy, high costs, and limited trust among patients and caregivers. Safe and effective integration requires robust data protection, rigorous validation, bias mitigation strategies, interdisciplinary collaboration, clinical integration, and continuous ethical oversight. When responsibly governed, AI holds substantial promise for advancing personalized, equitable, and data-driven palliative care.
Velibor Božić (Mon,) studied this question.