Trust boundary is a term in computer science and security used to describe a boundary where program data or execution changes its level of "trust". The term refers to any distinct boundary within which a system trusts all sub-systems (including data). An example of an execution trust boundary would be where an application attains an increased privilege level (such as root). A data trust boundary is a point where data comes from an untrusted source. For example, user input or a network socket
A "trust boundary violation" refers to a vulnerability where computer software trusts data that has not been validated before crossing a boundary.