Apple has reportedly asked Google to explore hosting servers to run a future, AI-powered version of Siri, according to a report by The Information on Monday. The move signals a deepening collaboration between the two tech giants, and potentially a significant shift in Apple’s approach to cloud infrastructure as it races to enhance its artificial intelligence capabilities.
The request centers on utilizing Google’s data centers to operate a Gemini-powered Siri, a version of the voice assistant that Apple announced in January would be bolstered by Google’s AI models. While the initial announcement detailed the integration of Gemini models into Apple Intelligence features, including a more personalized Siri, it did not specify where the processing would occur. This latest report suggests Apple is considering offloading some of the computational burden to Google’s cloud infrastructure.
Apple’s consideration of Google’s servers comes amid a historical reluctance to significantly expand its own cloud infrastructure, a strategy that has reportedly led to internal challenges and the departure of key personnel. The Information’s reporting details how Apple’s focus has traditionally remained on hardware and consumer-facing features, resulting in underinvestment in the cloud technologies needed to support increasingly demanding AI applications. Former Apple executive Patrick Gates, who pioneered the use of Apple silicon in data centers, left the company in part due to this prioritization, according to the report.
Currently, Apple relies on its “Private Cloud Compute” system, utilizing Apple silicon chips, for more complex AI tasks. However, utilization of this system remains low, averaging just 10% capacity, with some servers still awaiting installation. This underutilization occurred even as Apple’s internal AI infrastructure was described as “beginning to decay” and older, Nvidia-powered servers were being decommissioned. The combination of limited internal capacity and financial pressures has prompted Apple to increasingly consider third-party providers like Amazon.
For years, Apple maintained a policy of restricting its AI engineers from utilizing Google’s cloud technologies due to privacy concerns. Apple software chief Craig Federighi repeatedly vetoed Google Cloud as an option for its AI computing needs. However, Google reportedly made changes to its security systems in 2023 that addressed Apple’s privacy concerns, paving the way for the current exploration of server hosting arrangements.
The potential reliance on Google’s infrastructure arrives as Apple prepares to release its own conversational Siri model, anticipated with iOS 19. Apple had previously allowed Siri requests to be directed to OpenAI’s ChatGPT through Apple Intelligence, and code updates in February 2025 suggested the possibility of integrating Google’s Gemini as another option, potentially arriving in iOS 18.4 or a later update. The release of Gemini 2.0 models, including a new reasoning model, may further incentivize Apple to leverage Google’s AI capabilities, particularly for users of newer iPhone models like the iPhone 15 Pro, iPhone 16, iPhone 16 Pro, and iPhone 16e.
As of Tuesday, neither Apple nor Google have publicly commented on the specifics of the server arrangement discussions.
