After years of touting itself as the company that didn’t access its users data, Apple has finally acknowledged that it really does need at least some collective understanding.
Apple won’t collect your data for its AI services unless you let it
“Differential privacy” is Apple’s way of collecting data without being totally creepy.
The Verge
The company announced at last week’s developer conference that, starting with iOS 10, it will collect a range of new information as it seeks to make Siri and the iPhone better at predicting the information its owner might want at a given time.
Apple is looking to thread a fine needle, gaining access to the data it needs to make its servers smarter while also protecting user privacy. It’s doing so by employing a concept known as differential privacy.
However, the company was initially short on details on just what data it will be collecting and how. Here are a few things we’ve clarified over the past few days.
- Differential data is making its debut with iOS 10 and Apple says it has not yet been collecting such data.
- The decision to allow Apple use of data will be up to the user and require their opt-in consent.
- Apple says it is not using iOS users’ cloud-stored photos to power the image recognition features in iOS 10, instead relying on other data sets to train its algorithms. (Apple hasn’t said what data it is using for that, other than to make clear it is not using its users photos.)
As for what data is being collected, Apple says that differential privacy will initially be limited to four specific use cases: New words that users add to their local dictionaries, emojis typed by the user (so that Apple can suggest emoji replacements), deep links used inside apps (provided they are marked for public indexing) and lookup hints within notes.
Apple will also continue to do a lot of its predictive work on the device, something it started with the proactive features in iOS 9. This work doesn’t tap the cloud for analysis, nor is the data shared using differential privacy.
For example, iOS looks within mail and messages to find potential calendar items or contacts. Apple also looks at phone activity to suggest the apps that a user might want at a particular time as well as suggesting frequently used locations as options for calendar events.
With artificial intelligence increasingly taking center stage, many outsiders had wondered how Apple would keep pace. The company is clearly trying to stake a middle ground in which it can keep its reputation for privacy without entirely giving up on the kind of know-how that comes from a collective understanding of what users are doing.
While Apple is clearly pitching this as a just-right balance, it runs the risk of losing some of its privacy points while still not getting the kind of data it needs to truly rival Google and Facebook in the machine intelligence game.
This article originally appeared on Recode.net.
