Apple has begun giving secret “trust scores” to users based on how they use their devices.
The firm said that it has begun tracking the number of phone calls and emails sent and received from iPhones and Apple TVs to create a measure of trustworthiness.
The scores could help Apple tackle fraud by tracking fake reviews and spam App Store accounts. However, Apple has refused to reveal the exact reasons why it is collecting the data.
Apple’s new policy was shown in the legal agreement for its iOS 12 software update, which was released on Monday.
The agreement reads: “To help identify and prevent fraud, information about how you use your device, including the approximate number of phone calls or emails you send and receive, will be used to compute a device trust score when you attempt a purchase.”
“The submissions are designed so Apple cannot learn the real values on your device. The scores are stored for a fixed time on our servers.”
This is the first time that Apple has told its customers about assigning trust scores, and it is unclear how the number of phone calls or emails will help determine them.
The trust scores themselves are calculated on Apple devices, and the score sent to the company cannot be used to identify an individual user or whom they have contacted. The scores are not used to target adverts.
Tracking customer behaviour is a strange development for Apple, which has publicly criticised the user-tracking business models of other large US technology businesses such as Google and Facebook.
In June, Apple chief executive Tim Cook said that tracking was “totally out of control.”
“I think most people are not aware of who is tracking them, how much they're being tracked and the large amounts of detailed data that are out there about them”, Mr Cook told CNN.
Apple doesn’t store the contents of emails and phone calls, and instead told customers that it only tracks the volume of interactions.
Facebook also recently announced that it had begun assigning trust scores to its users. The company said last month that it gives all users a numerical value, which is based on which news stories users report as untrue, and whether those stories were actually false.