AI in palliative care: a scoping review of foundational gaps and future directions for responsible innovation

This scoping review systematically mapped the landscape of AI applications in palliative and hospice care, screening 4,747 records and identifying 125 studies meeting inclusion criteria. The findings paint a picture of a field still in its early stages: over 86% of studies were retrospective proof-of-concept designs, AI applications were concentrated heavily in mortality prediction and cancer populations, and transparency was strikingly limited, with only 15% of studies sharing code, 11% providing data access, and none adhering to AI-specific reporting guidelines. Ethical frameworks for evaluation were notably absent across the literature.

These findings raise important concerns about the readiness of AI tools for deployment in one of medicine's most sensitive settings. End-of-life care demands a particularly high standard of reliability, fairness, and transparency, yet the current evidence base falls short on external validation, cross-site testing, and open science practices. The review calls on researchers and developers to prioritize rigorous validation, broader patient representation, and greater reproducibility to ensure that AI tools in palliative care are not only effective but genuinely trustworthy before reaching clinical use.