iPad Dictation Prompts Apple Data Retention Questions

There was a privacy scare recently when the iOS app Path was discovered to be uploading contact and user information to their servers without asking the owners permission. Path quickly rectified the situation by pushing out an update to their app, and Apple also included new warnings in Mountain Lion Developer Preview 2, making it clear when an app was trying to access your contact information. So what about the Dictation feature found on the new iPad?ZDNet decided to take a closer look at the Dictation feature including Apple's terms of use and privacy disclosure. While it shouldn't come as any surprise that your voice is being sent to Apple's servers in order to determine what is being said, there are some notable pieces of information that can be gleaned from Apple's privacy form.

First, its not just your voice that floats up to the cloud to be processed: other information about you is also uploaded, such as the names of your contacts, and song names. That's so that the Dictation service can understand what it is you're asking for, as well as gaining context.

What about if you decide to turn off Dictation? Is your information stored at all? Yes and no. Apple say that user data, like contacts and song names, will be deleted from their servers, as well as recent voice samples. Apparently not everything is though:

Older voice input data that has been disassociated from you may be retained for a period of time to generally improve Dictation and other Apple products and services. This voice input data may include audio files and transcripts of what you said and related diagnostic data, such as hardware and operating system specifications and performance statistics.

The question then is what does Apple consider to be "disassociated" data, and how long do they keep it for? We're waiting for a response from Apple.