Silicon Valley leaders have touted artificial intelligence as a near-term path to curing cancer, but clinicians and researchers say biology’s complexity and regulatory hurdles mean progress will be incremental, not absolute. While AI is improving diagnostics—spotting faint signals in imaging, guiding brain-tumor workflows, and enabling faster, more tailored care—experts caution that cancer is not a single disease and lacks a one-size-fits-all cure. Studies show machine learning can flag pancreatic cancer earlier and speed surgical decision-making, yet most AI tools remain in trials amid unclear approval pathways, especially in Europe. Despite promises that more computing power will “solve” disease, researchers argue discoveries still hinge on high-quality data, careful validation, and human experimentation. The likely payoff: more efficient, equitable, and personalized oncology—not a universal cancer cure.
Related articles:
International evaluation of an AI system for breast cancer screening
End-to-end lung cancer screening with deep learning on low-dose CT
DNA methylation–based classification of central nervous system tumours
Highly accurate protein structure prediction with AlphaFold
WHO: Ethics and governance of artificial intelligence for health
