Apple offers more information than ever
The site also offers a selection of understandable white papers that explain how privacy controls work in Safari, Location Services, Photos and Sign-in With Apple. These contain a large amount of information on Apple and its services.
Hey Siri, spell 'debacle'
Apple likes to update users with more information concerning its ongoing attempts to protect user privacy annually, but the update is particularly significant this year following the debacle around Siri grading.
Since that news broke, Apple has taken steps to make the Siri grading process more transparent – and to give users more control. Not only can you now avoid the process completely, but you can also delete any recordings Siri’s system may have on you. The company has also taken the grading process in-house.
Apple’s decision to protect user privacy isn’t just altruistic: It recognizes that each time you pick up a smartphone, make a call, check a message or visit anywhere, the sensors inside the device can track your action. In short, smartphones gather a huge quantity of information about you.
Apple says it has designed its systems so such information stays on your device and data is not shared with the company (beyond what is required to make systems work – and even then it is encrypted).
[ Further reading: The wireless road warrior’s essential guide ]
That’s unique in the industry, as server-based smartphone platforms (you know who they are) do gather this information in order to make the OS (and their revenue models) work.
Your iPhone uses this kind of information on a device level, but only to assess things like human/not human requests, as are typical of developers and other service providers who ask for bot/not bot data.
(Apple calls this its Real User Indicator.)
This kind of system is what helps prevent large-scale advertising click fraud, for example – if a device tells the service it thinks it is being run by a bot, then that click can be ignored.
Apple’s new website goes into plenty of detail concerning how you can use its platforms to protect privacy, including a plea to the one-in-four iPhone users who have not yet done so to enable two-factor authentication.
The page also explains the many privacy improvements in iOS 13, including much tighter controls over location information. For example, when you install a new app you now have the opportunity to share location data just once in order to try the app and make it work.
Apple’s system now provides users with highly visual reminders of what information specific apps gather about them as a way to jog us all into reviewing which apps gather what data from time to time.
Inside the white papers
Apple’s new white papers are relatively approachable documents for any iOS user. They carry a wealth of detail concerning Apple’s products and approach, including a few items I’d not been aware of before. Did you know that over a trillion photos and videos are captured every year on iPhones?
I’ve gathered one interesting data point from each of the white papers, but please do Tweet me with any that catch your eye (as I’d urge you to read them):
Location Services White Paper:
“Apple Maps shows your current location using a blue marker. Because the precision with which your location can be determined is limited, a blue halo will appear around this marker. The size of the halo shows approximately how precisely your location can be determined: the smaller the halo, the greater the precision.”
Photos White Paper:
“In order for the Days view to highlight a user’s best shots, photos that are considered clutter are shown in the All Photos view and intelligently hidden in the Days view.
"Days automatically hides:
-- Screen recordings
-- Blurry photos
-- Photos with bad framing and lighting
Photos also identifies specific objects likely considered clutter and hides them in the Days view. They include photos of:
-- Credit cards.”
Sign in with Apple white paper
“Apple has gone to great lengths to ensure the [Real User Indicator] is calculated in a privacy-preserving manner. First, on-device machine learning (ML) is employed to measure if the device the account is originating from is being used in a way that’s consistent with ordinary, everyday behavior such as moving from place to place, sending messages, receiving emails, or taking photos. This analysis yields a tamper-proof numerical score that is sent to Apple indicating a level of confidence that the device is being used by a real person. The score cannot be reverse-engineered by Apple to reveal any personal information, and none of the specific inputs to the ML models ever leaves the user’s device.
"The device score is then combined with select information on recent account activity. The sum of this information is abstracted into the binary Real User indicator that is passed to the developer at account setup time. Developers can incorporate this information into any existing anti-fraud systems they currently use to help make determinations about how to handle a new customer.”
In other words, your device figures out whether you are a real person or not. Developers/service providers can then choose not to demand additional verification, or to do so if you might be a bot.
Safari White Paper
“Safari now includes the ability to offer Private Click Measurement, an innovative way of doing ad click measurement that prevents cross-site tracking but still enables advertisers to measure the effectiveness of web campaigns.
“It is built into the browser itself and runs on device, which means that neither advertisers, merchants, nor Apple can see what ads are clicked or which purchases are made. This solution avoids placing trust in any of the parties involved—the ad network, the merchant, or other intermediaries—so none of them are able to track users as they click on ads and make purchases in Safari…. “Apple has proposed Private Click Measurement as a new web standard to the World Wide Web Consortium.”