The article proposes a model for steganographic embedding of data into a stream of messages using error-correcting coding. Embedded data is disguised as interference in the information channel. The inline message extraction is based on the error correcting code detection procedure for corrupted bits. The robustness of the scheme is based on the inability to obtain complete information about the embedding by the analyst. The model is universal and can be implemented in any network protocol that uses error-correcting codes.