Issue Reimbursement evidence, cost gaps and ethics frameworks for AI-assisted embryo selection

Summary

AI embryo assessment software as a medical device is expected to enter first as a non-reimbursed service and expand step by step through institutional schemes. Insurance coverage requires clinical evidence centred on live births, yet most current studies rely on surrogate indicators such as pregnancy and implantation. Reliability therefore calls for randomised controlled trials, live-birth outcome measures, external data validation and continuous evaluation of real-world performance and post-market safety. Non-invasive genetic estimation models that claim to replace or complement preimplantation aneuploidy testing need an ethical framework on decision responsibility, explainability and protection against misdiagnosis, alongside updated hospital ethics committee and clinical guidelines. Because early costs and non-covered services dominate, KISTEP also warns that access may concentrate among high-income groups.