The session examines one of the most practical barriers to healthcare AI adoption: how health systems assess and approve AI vendors without either compromising security or allowing legacy review processes to become an indefinite barrier to implementation. I envision bringing together a representative from CORL/Meditology, an AI vendor that directly works with these processes such as Jason Campbell from Rad AI, a health-system leader who directly oversees vendor security or third-party risk reviews (ideally Advent given the proximity), and Karen Mandelbaum of Epstein Becker Green, who previously advised the CMS CIO and CISO on security and privacy policy. The discussion focuses on what is genuinely different about AI security review, where approvals get stuck, what evidence health systems should require, and how responsibility should be managed after deployment.