Published on March 3, 2026 at 10:00
Comment
0
Siri Icone Logo © 01net

The future Siri, which was supposed to start appearing in iOS 26.4, relies on Google's AI models. But it's not the only web giant's technology that the apple company needs. It should also entrust the hosting of the new assistant to Google servers, but configured according to Apple's data protection standards.

In June 2024, Apple presented the new Siri, an AI-powered assistant capable of searching through the user's data and apps to answer questions in a much more personal and precise way.

Alas, more than 18 months later, still nothing! We had hoped that iOS 26.4, currently in beta, would finally bring the first features of this Siri 2.0, but so far it's still dead in the water.

Even Siri can no longer do without Google

In the meantime, Apple has partnered with Google so that Siri can leverage the Gemini AI models, which the web giant has agreed to modify and customize to meet the needs of this new client. Nevertheless, this remains an admission of Cupertino's shortcomings in the field of AI.

Apple could also rely on another Google technology, or rather another infrastructure. The site The Information confirms a rumor that has been circulating for weeks: the new Siri would be hosted in Google's data centers... while still respecting the privacy standards imposed by Apple. Who will have to show some patience to get this pill down to its most privacy-conscious users.

Apple has always prided itself on mastering the essential building blocks of its products, from chip design to the operating system. However, in the field of cloud computing and generative AI, the company is increasingly relying on external actors. It already uses Google Cloud for certain tasks, such as online storage or training its internal models.

This increased reliance on Google comes as Apple's in-house cloud infrastructure shows its limitations. Apple highlights its Private Cloud Compute system, intended to handle the most complex AI queries on servers equipped with internally designed chips. Simpler queries, however, would still be processed locally on the iPhone.

According to the site, this cloud capacity would be used on average only 10%. Some servers intended for AI have not even been installed yet. More concerning still, the Apple chips deployed in these data centers would not have been designed to run very large models such as those of Gemini. Finally, software updates would also be slower in these facilities compared to other infrastructures.

In addition, for years, Apple's financial leadership would have viewed the cloud as a cost rather than a strategic investment. Apple's siloed operations — with each team working in its own area — also do not help in consolidating computing hosting capabilities in the cloud. It becomes clearer then why it is so difficult to launch this new Siri, which combines several specializations: generative AI, massive data processing, and strict privacy requirements.

👉🏻 Follow real-time tech news: add 01net to your Google sources, and subscribe to our WhatsApp channel.

Source : The Information

This increased reliance on Google comes as Apple's in-house cloud infrastructure shows its limitations. Apple highlights its Private Cloud Compute system, intended to handle the most complex AI queries on servers equipped with internally designed chips. Simpler queries, however, would still be processed locally on the iPhone.

Source : The Information

This increased reliance on Google comes as Apple's in-house cloud infrastructure shows its limitations. Apple highlights its Private Cloud Compute system, intended to handle the most complex AI queries on servers equipped with internally designed chips. Simpler queries, however, would still be processed locally on the iPhone.

Source : The Information