Impact of Artificial Intelligence on the Care of Terminally Ill Patients

Introduction: In recent decades, demographic aging has led to an inversion of the population pyramid, with a marked increase in the proportion of older adults. This shift has been accompanied by a higher prevalence of chronic and life-limiting diseases, while there have also been significant technical and scientific advances. However, these developments have not been matched by a proportional expansion of healthcare human resources, including in palliative care (PC). Consequently, healthcare systems face increasing pressure, particularly in the provision of end-of-life care. Artificial intelligence (AI) has emerged as a promising tool to support and improve healthcare delivery. Objective: This study aims to review the literature on the impact of AI on palliative care, with particular emphasis on its clinical applications and ethical implications in end-of-life care. Methods: A narrative review was conducted using a structured search of PUBMED, CINAHL and Web of Science databases, covering publications from the last ten years (2015-2025). Search terms included combinations of "artificial intelligence", "machine learning", "palliative care", "end-of-life care", and "ethics". Articles were included if they addressed clinical applications, implementation challenges or ethical aspects of AI in PC. Reference lists of selected articles were screened to identify additional relevant studies. The findings were analyzed and synthesized thematically into key domains of application and ethical concern. Results: The literature suggests that AI is currently a promising tool in PC, particularly in prognostication, symptom assessment, clinical decision support, and communication. These applications may represent a paradigm shift compared to conventional approaches. However, it is important not to forget that patients in PC need much more than algorithmic decision trees. Thus, current evidence is largely exploratory, with limited real-world validation. Empathetic emotional support, physical comfort and compassion are things that artificial intelligence cannot provide. AI does not replace humans in interpersonal relationships and dignity; it only complements them. Conclusions: AI-based technologies hold significant potential to address contemporary challenges in PC, including inequitable access, workforce strain, and the need for more efficient service delivery. Nevertheless, their implementation raises substantial ethical concerns related to autonomy, transparency, data governance, and the preservation of human dignity. AI should therefore be understood as a complementary tool that supports-but does not replace-the human dimension of PC.