Apple Collecting, Storing Customer Data for 18 Months

Apple has admitted to retaining private personal data collected by services Siri and Dictation for a period of 18 months.

The disclosure came after one civil liberties group complained that Apple is not doing enough to inform its customers of their privacy rights. In response, Apple has revealed the kind of customer information that is collected and stored.

In a recent interview, Apple spokeswoman Trudy Muller explained how her company performs their data collection.

Siri (available on iOS devices) and Dictation (available on both iOS and OS X devices) send speech-to-text messages wirelessly to Apple servers.

All About Improving Service, Apple Claims

A random number is then assigned to anonymize the individual. Muller noted that the transmitted information excludes email addresses, phone numbers and Apple IDs. (Source: mobilenapps.com)

After six months, the data becomes "disassociated" from the random number and Apple retains the information only in an effort to "generally improve [Siri / Dictation] and other Apple products and services".

Apple retains these "disassociated files" for up to 18 months on their servers.

Muller insists Apple collects data for the sole purpose of improving its services, adding that his "company takes privacy very seriously."

IBM Bans Apple Voice Devices on Corporate Network

Up until now, Apple's voice-driven privacy standards have been rather vague; however, the official policy does warn that personal data may be recorded (although it does not state exactly how the company does this).

Nevertheless, the warning has been enough for some companies to forbid Apple devices from entering into their place of business. In 2012, IBM banned Siri-related messages from appearing on their corporate network, fearing "spoken queries might be stored somewhere."

Nicole Ozer, technology and civil liberties policy director at the ACLU of Northern California, speculated as to the kind of information Apple can (and likely has) retained: "first names, nicknames, relationships via address book contacts and other user data." (Source: zdnet.com)

Google Guilty of Similar Practice

Google has taken a similar approach to storing customer data.

As with Apple, Google acknowledges that this practice helps improve their overall speech recognition service in the long run. But unlike Apple, Google assigns no random number to the individual so there is "no way of telling who spoke a particular query."

Still, for many critics the fact that words uttered in confidentiality are being read and stored remains rather unnerving.