Google as a cloud platform for Siri? New statements raise questions

The question of where the new, Google Gemini-powered version of Siri is technically operated remains unanswered. Despite numerous public statements from Apple and Google executives, there is still no clear confirmation of the underlying infrastructure. Instead of providing clarity, recent statements from Google management have further complicated the situation while simultaneously offering new clues.

Since Apple officially confirmed that Google's Gemini technology will enable new Siri features, the focus has shifted from functionality to data privacy. Apple has emphasized for years that AI processing preferably takes place on the device or via its own private cloud computing infrastructure. Google, on the other hand, is primarily positioned as a cloud provider.

It is precisely at this interface that uncertainty arises: If Gemini powers Siri, the question inevitably arises whether user data is processed on Google servers or whether Apple retains full control over the infrastructure.

Apple's first statement on the collaboration with Google

Following the announcement of the partnership, Apple stated that it had chosen Google's technology after careful evaluation, as it provided the most powerful foundation for its Apple Foundation Models. At the same time, the company pointed out that its own Foundation Models run either directly on the device or via private cloud computing.

These statements gave the impression that the Gemini integration would also take place entirely within Apple's infrastructure. However, concrete technical details or an explicit distinction between existing Apple models and the new Gemini-based technology were lacking.

Bloomberg presents an alternative account

A few days later, Bloomberg reported that Apple and Google were discussing hosting the Gemini chatbot directly on Google servers. These servers would reportedly be powered by Google's own Tensor Processing Units. According to the report, only a short-term Siri update would run through Apple's private cloud computing, while more advanced AI functions could be outsourced to Google's infrastructure.

This portrayal challenged previous assumptions and intensified the discussion about Google's actual role.

Tim Cook's statements remain general

During the announcement of the quarterly results, Tim Cook commented on the decision to partner with Google. He explained that Google's AI technology provided the best foundation for Apple's Foundation Models and that the collaboration would enable new innovations. At the same time, he reiterated that Apple would continue to work on-device and private cloud computing and maintain its data privacy standards.

It was striking that Cook made no direct connection between these statements and the Gemini integration. Financial or infrastructural aspects of the agreement were also left unmentioned.

Google clearly states its own role

During Alphabet's Q4 2025 results conference call, Sundar Pichai said Google is working with Apple, its preferred cloud provider, to develop the next generation of Apple Foundation Models based on Gemini technology.

Shortly afterwards, Google's Chief Business Officer repeated this statement almost verbatim. Both formulations made it clear publicly for the first time that Google not only supplies technology, but is actively involved in its development as a cloud provider.

This choice of words suggests that at least part of the new Foundation Models, and therefore very likely also the Gemini-based Siri, will be operated on Google servers.

New interpretation of previous Apple statements

These statements offer a new interpretation of previous communication:

Apple's references to on-device processing and private cloud compute appear to refer to existing Apple intelligence features.
The statement that Google's technology forms the basis for the Apple Foundation Models seems to refer to a new generation of these models.
Google is explicitly named as the preferred cloud provider, without Apple clarifying which functions are affected.

This view is supported by a renewed assessment from Bloomberg journalist Mark Gurman. He reiterated that Apple may be treating Siri and Apple Intelligence as separate systems. According to him, Siri could run on Google Cloud this fall, while other AI functions would remain on Apple's own infrastructure.

Why Apple and Google are leaving the core question unanswered

So far, neither Apple nor Google has definitively confirmed where the Gemini-based Siri is technically operated. All public statements remain vague and avoid concrete details. This reticence suggests that the infrastructure question is being deliberately left open.

The repeated designation of Google as the preferred cloud provider suggests that the new Siri runs at least partially on Google's servers. Whether and how Apple will address this fact regarding data protection and data processing during the official unveiling of the Gemini-based Siri remains to be seen. What is clear, however, is that the collaboration between Apple and Google runs deeper, both technically and strategically, than previous statements initially suggested. (Image: Shutterstock / Tada Images)