The article discusses the development of a system for identifying the channel of information leakage from a document photo taken from a computer screen via digital watermarks integrated into the image on the monitor. The article also presents the results of an experimental study to prove the properties of the algorithm and its resistance to deliberate attacks.