Secure messaging lets you create policies for storing mail locally rather than delivering it, and requiring recipients to log on over HTTPS to securely view their messages. It works in conjunction with outbound filtering.
In use you set up CanIt policies to trigger on key content in the header or body of outbound messages. When a secure message triggers on a key the message is kept in CanIt and not delivered to the recipient. The recipient receives notification that they have a secure message, along with a link and details about how they access it over HTTPS. If this is the first time they have received a secure message from this site they will be instructed on how to connect and create an account. This is only necessary the first time.
Setup is described in Chapter 11 of the most recent User Guide.