An audio/video monitoring system is disclosed, comprising a combination of one or more set-top boxes and application servers in a service provider's network. The set-top boxes and applications services are part of a multiple services operator's (MSO) network, such as that of a cable company. Cameras and/or microphones which are directly connected to a set-top box are the sources of media signals, which the set-top box transmits to an application server. The application server continually receives the media signals from one or more set-top boxes, and determines which subset of the received media signals to forward to a selected output device, such as a television connected to another set-top box or a cellular telephone. The application server then forwards media signals of interest to the output device, such as when motion or a loud noise is detected in a media signal received from the set-top box or boxes.