Generative artificial intelligence in oncology patient counselling: Transforming oncology pharmacy practice through human–AI collaborative care

Abstract

Objective: To critically evaluate the current evidence regarding the application of generative artificial intelligence (GenAI) in oncology patient counselling, with particular emphasis on oncology pharmacy practice, and to distinguish direct oncology evidence from indirect evidence derived from general healthcare and other digital-health technologies. Data sources: PubMed/MEDLINE and Web of Science were searched for relevant literature published from January 2020 through August 2026, with the final search conducted in August 2026. Search terms included combinations of terms related to generative artificial intelligence, large language models, ChatGPT, GPT, Med-PaLM, oncology, cancer, patient counselling, patient education, medication counselling, medication adherence, symptom monitoring, toxicity monitoring, telepharmacy, and digital health. Reference lists of relevant publications were also examined to identify additional relevant literature. Data summary: The identified literature demonstrated substantial heterogeneity in artificial intelligence technologies, clinical settings, populations, and outcomes. Evidence directly evaluating GenAI in oncology patient education and counselling remains limited but suggests potential applications in patient education, healthcare communication, information simplification, and selected supportive-care functions. Evidence from general healthcare populations and from conventional chatbots, telehealth, remote monitoring, and other non-generative digital-health technologies provides contextual but indirect support. Evidence for GenAI-specific improvements in medication adherence, toxicity recognition, hospitalization outcomes, accessibility, workflow efficiency, and continuity of care remains insufficient. The review also identified important concerns regarding hallucinations, accuracy, bias, privacy, accountability, and the need for professional oversight. A pharmacist-supervised human–AI collaborative framework is proposed as a conceptual model based on the reviewed evidence, existing AI governance principles, and oncology-pharmacy considerations. Conclusions: GenAI represents a potentially useful adjunct to oncology patient counselling, particularly for patient education and communication. However, current evidence does not establish its clinical effectiveness across major oncology-pharmacy outcomes. Findings from non-generative digital-health technologies should not be interpreted as direct evidence for GenAI. Prospective oncology-specific studies are needed to evaluate clinical effectiveness, patient-centred outcomes, medication-related outcomes, safety, usability, and implementation feasibility. Until stronger evidence is available, GenAI should complement rather than replace oncology pharmacist-led counselling and clinical decision-making.