In some examples, a trust controller generates a first value and send the first value to a target controller of a subsystem, and generates a first verification value based on the first value and a known good code image for the target controller. The trust controller receives a second verification value from the target controller, the second verification value based on the first value and a code image to be executed at the target controller. The trust controller determines whether the code image to be executed at the target controller is compromised based on the first verification value and the second verification value.