Apple has asked Google to investigate running Siri on its servers as the tech giant prepares for a more powerful, Gemini-based version of its voice assistant. Currently, Apple's Private Cloud Compute system, which runs on Apple silicon chips, uses only about 10% of its capacity on average, with servers still sitting in warehouses. This underutilization could rapidly change when the next-generation Siri launches.

Apple's cloud infrastructure challenges stem from a long-standing cultural focus on hardware over cloud services, leading to the departure of key cloud experts like Patrick Gates. The company's AI infrastructure was described as "beginning to decay" as it decommissioned old Nvidia-powered servers while facing financial pressure. For years, Apple banned its engineers from using Google's cloud technologies due to privacy concerns, with software chief Craig Federighi repeatedly vetoing Google Cloud as an option.

In 2023, Google made security changes that satisfied Apple's privacy requirements, prompting the company to adopt Google's cloud infrastructure for AI workloads. The timing of Apple's request is critical as Private Cloud Compute servers, designed for consumer devices rather than AI workflows, struggle to handle large models like Gemini efficiently.