Apple May Use Google Servers for AI Siri
Apple is reportedly considering using Google's cloud servers to store data for its upcoming AI-powered Siri upgrade. The move highlights the massive infrastructure challenge Apple faces in competing with other AI assistants and its need to balance privacy with performance at scale.
This potential partnership is part of a broader initiative, internally codenamed "Project Greymatter," to embed generative AI across Apple's core apps like Safari, Photos, and Notes. The strategy employs a hybrid model: simpler AI tasks are handled on the device itself, while more complex requests are sent to the cloud. Apple refers to its cloud AI infrastructure as Private Cloud Compute, which is designed to handle more intensive AI tasks securely. The company has emphasized that its servers, running on Apple Silicon, do not store user data. This move with Google would be an expansion of this hybrid approach, leveraging a competitor's massive infrastructure to accelerate capabilities. The collaboration with Google is expected to center on using Google's Gemini AI models to power a significantly upgraded, more conversational version of Siri. This partnership is seen by analysts as a pragmatic way for Apple to accelerate its AI development and reduce time-to-market, rather than building every component from the ground up. While partnering with a rival, Apple is also making substantial investments in its own AI infrastructure. The company has announced a four-year, $500 billion investment in the U.S., which includes building a new server manufacturing facility in Texas and expanding its data centers. However, its AI-related capital expenditures of about $12.72 billion in fiscal 2025 remain significantly lower than competitors like Google and Microsoft. The discussions reportedly involve Google setting up dedicated servers that adhere to Apple's stringent privacy standards. One possibility is deploying Google's servers within Apple's own data centers, allowing Google's AI systems to operate inside Apple-controlled facilities. This move highlights the immense computational power required for generative AI, which has pushed even a giant like Apple to seek external partners. Google Cloud already provides some services for Apple, including data storage and training for some of its in-house AI models.
