Approaches presented herein enable visualization of audio announcements using augmented reality (AR). Specifically, an audio system captures a set of broadcast audio announcements or messages at an originating location associated with a user as a set of audio files, which a speech-to-text transcriber transcribes to text. A natural language processing tool analyzes the transcribed text to determine, for each announcement, whether the announcement is relevant to the user. Relevant announcements are forwarded to a mobile device of the user. The user may access, in an AR environment, the set of announcements via a set of icons to display text of an announcement or to play audio of an announcement.