A new study looks into Apple’s use of “differential privacy,” which inserts random noise into an individual user’s information to help protect user data from being exposed by a third party, and contends the company may be supporting more data mining than its public promise implies, Wired reports. Researchers found that Apple’s use of the practice uploads more specific data than the typical differential privacy researcher might consider private, adding that Apple keeps both the code and the “privacy loss parameter,” or the amount a data collector is willing to sacrifice for the sake of protecting its users’ privacy, a secret. In response, Apple states that its data collection is purely opt-in, highlighting that it prompts users to share “diagnostics and usage” information with the company when its operating systems first load.
Full Story