Apple Intelligence is playing privacy game as a killer feature

Like Apple of Worldwide The Keynote of the developer conference ended on Monday, the observers of the market could not help but notice that the price of the company’s shares was falling, perhaps a reaction to the relatively low Apple approach to incorporate the IA compared to most of its competitors. However, the characteristics and updates based on Apple intelligence were abundant and, although some are fed using the cloud platform on privacy and the safety of the company known as private Caltal cloud, many work locally on devices qualified for Apple intelligence.

The screening function of Apple’s new messages automatically moves the texts from telephone numbers and account with which you never interacted first in a “Sennder unknown” folder. The functionality automatically detects messages sensitive to time as access codes or food delivery updates and will still deliver them to your mailbox, but also scan for messages that seem to be scams and put them in a separate spam folder. All this sorting is performed locally using Apple Intelligence. In the same way, the expanded call screening function will automatically and locally resume the unreliable phone calls, ask for details on the caller and will transcribe the answers so that you can decide if you want to answer the call. Live translation also adds a linguistic translation in real time to calls and messages using local processing.

From the point of view of privacy, local processing is the Gold Standard for artificial intelligence features. The data never leave your device, which means that there is no risk that it can end in some unintentional way due to a journey through the cloud. And new features such as spam and “sentder unknown” for messages, calls for non -reliable telephone numbers and live translation tools seemed to be designed with a strategy of using privacy as a differentiation in a field already crowded.

In addition to being suitable for privacy, local processing has other advantages such as allowing services based on artificial intelligence to be available offline and accelerate certain activities since the data must not be sent to the cloud, processed and therefore sent to a device. If the artificial intelligence features are widely available and accessible, however, most companies are bound in an attempt to take into account the old low -end devices that many of their customers probably use that they may not be able to manage local artificial intelligence. Apple needs less to be inclusive, however, because it produces both hardware and software and has already imposed limits that Apple Intelligence can only perform on recent devices models.

There are also other limitations for Apple Intelligence and the company offers opt-in additions with some third-party artificial intelligence services to expand functionality. For Open’s Chatgpt, for example, users must activate the integration and Apple Services will push the user to confirm every time they go to send a chatgpt querg. In addition, users can choose to access an account chatgpt, in which case their query will be subject to the normal Openi policies, or they can use chatgpt without accessing. In this scenario, Apple claims that it does not connect an Apple ID or other iP identification to the users of users.

Apple has invested widely to develop a private cloud calculation to maintain strong safety and privacy guarantees for the era processing in the cloud. Other companies have even started creating similar cloud schemes for the safe for products and services that specifically center privacy as a crucial feature. But the fact that Apple still distributes local processing for new features, when possible, may indicate that privacy is not only an intellectual priority in the company’s approach to the AI, it could be a corporate strategy.