A set of developers and researchers highlighted advances in non-invasive, data-driven cancer diagnostics in a multicenter setting, including AI approaches aimed at reducing unnecessary procedures. One example: an imaging-focused model for breast microcalcifications reported high accuracy for malignancy prediction in BI-RADS 4 cases. While the reported results are positioned for clinical workflow integration, the key industry angle is how image-based risk stratification could change downstream care pathways by identifying patients most likely to benefit from biopsy. At the same time, oncology AI remains under pressure to demonstrate external validation and robustness across diverse centers—an issue that continues to shape payers’ and clinicians’ confidence in deploying new tools beyond retrospective datasets.