Apple in Talks to Host Siri on Google Servers

Apple is in advanced discussions with Google to host its upgraded Siri virtual assistant on servers within Google's data centers, as the company grapples with limitations in its own cloud infrastructure. The move, reported by industry sources, would support the rollout of a more advanced, conversational version of Siri powered by Google's Gemini AI models, expected to launch in the coming months alongside updates to Apple's operating systems.

The collaboration builds on a multi-year agreement announced by the two companies earlier this year, under which Apple's foundational AI models for features like Apple Intelligence will leverage Gemini technology and Google Cloud. While Apple has emphasized that much of its AI processing will continue through its Private Cloud Compute system, designed to handle sensitive tasks on Apple silicon-based servers, that setup is currently operating at only about 10 percent capacity. Some hardware intended for Private Cloud Compute remains uninstalled, and the system, built on chips originally optimized for consumer devices, struggles with the intensive demands of large-scale AI workloads, such as multistep queries and extended conversations.