Genetic Counselors

The BRIDGE trial (Biesecker et al., JAMA Network Open 2024) is the most significant single piece of evidence for AI displacement in genetic counseling to date: a rigorously designed RCT with N=3,073 randomized to either an AI chatbot (with registered nurse oversight) or a board-certified genetic counselor for pre-test hereditary cancer counseling. The chatbot arm was declared statistically non-inferior on patient knowledge, decisional conflict, satisfaction, and testing uptake — the four primary outcomes that define the service's clinical value. This is not a proof-of-concept study; it is a published, peer-reviewed equivalence demonstration in a generalizable population. Health systems are actively citing BRIDGE as justification for deploying chatbot-first pre-test pathways.

AlphaMissense's classification of 71% of all possible human missense variants (89% sensitivity, 90% specificity on ClinVar validation set) represents a structural shift in the variant interpretation pipeline: the variant-of-uncertain-significance (VUS) rate, which was ~40–60% of variants reported in hereditary cancer panels as recently as 2020, is declining as AI classification resolves previously ambiguous variants. Simultaneously, commercial platforms (Fabric Genomics, Emedgene, Franklin by Genoox) have automated the ACMG five-tier classification workflow including evidence code application, database querying, and report generation — tasks that previously required 1–3 hours of laboratory genetic counselor or clinical geneticist time per variant.