The editorial of issue #962 of the Revue Médicale Suisse, published on May 13, 2026, explores the growing role of artificial intelligence in oncology, highlighting both its transformative potential and the challenges surrounding its integration into clinical practice.

Written by Pelagia Tsoutsou, Alfredo Addeo, Fernanda Herrera and Solange Peters, the editorial presents AI as an increasingly important tool across the oncology continuum—from early cancer detection and diagnostic pathology to precision medicine, radiotherapy optimization, and clinical decision support.

One of the most advanced applications of AI lies in early diagnosis, particularly through deep learning models applied to medical imaging such as CT scans, MRI, and mammography. In some settings, these systems already demonstrate diagnostic performances comparable to—or even exceeding—those of human experts by identifying subtle imaging patterns that may escape visual detection.

The editorial also highlights major advances in digital pathology, where AI-driven image analysis improves tumor classification, biomarker quantification, and diagnostic reproducibility, while reducing turnaround times.

Beyond diagnosis, AI is emerging as a key enabler of precision oncology. By integrating multi-omics data—including genomics, transcriptomics, epigenomics, and metabolomics—with clinical and radiological information, AI models can help predict treatment responses and identify novel therapeutic vulnerabilities, paving the way for increasingly individualized cancer care.

The authors devote particular attention to radiotherapy, a field especially suited to AI integration because of its reliance on large volumes of digital data. AI applications already support automated tumor segmentation, treatment planning, dose optimization, and adaptive radiotherapy approaches that dynamically adjust treatment according to anatomical changes during care.

At the same time, iimportant challenges remain. Data quality, representativeness, and standardization are essential to avoid algorithmic bias and unequal outcomes. The “black box” nature of some deep learning systems also raises concerns regarding interpretability, transparency, and clinical trust. Regulatory validation and clear demonstration of patient benefit will be necessary before widespread adoption.

Importantly, AI should not replace clinicians, nor the human dimension of care. Instead, it should be viewed as a tool that augments medical expertise and supports more precise, efficient, and patient-centered oncology.