Dictation traffic is the audio or speech data an application receives and converts into text, commands, or other structured content. It may include raw microphone input, transcribed text, and any metadata needed to route that input to a local client or cloud service. In security terms, it is a sensitive data path, not just a user-interface feature.
It matters because dictated speech can contain passwords, prompts, business details, or authentication material before the app has filtered it. If malware gains a foothold on the same device, it may try to redirect, intercept, or replay that traffic by changing local settings, hooking the input path, or tampering with the destination. Defenders reduce risk with least privilege, hardened endpoints, tamper-resistant configuration, outbound destination monitoring, and short-lived tokens that limit the value of any captured data.