Apple May Use Google Servers for AI Siri
In a significant potential partnership, Apple is reportedly considering using Google’s cloud infrastructure to store and process data for its upgraded, AI-powered Siri. The move highlights the massive computing scale required for modern AI, but also raises questions about data privacy and Apple's reliance on a chief rival.
This partnership is a key part of Apple's "Project Greymatter," a broad initiative to embed AI features across core apps in iOS 18. The collaboration would specifically leverage Google's powerful Gemini models to handle complex, server-based generative AI tasks. Apple's strategy is a hybrid one: most "Apple Intelligence" features will run directly on the device for privacy and speed. Only the most demanding queries that require massive computational power would be sent to the cloud, which is where Google's infrastructure would play a role. A high-stakes financial relationship between these rivals is not new. Google currently pays Apple billions of dollars each year to remain the default search engine on Safari, a long-standing deal that has attracted antitrust attention. To mitigate privacy concerns, Google would reportedly be required to adhere to Apple's stringent privacy standards. One potential arrangement involves Google setting up and running its servers directly inside Apple's own data centers, giving Apple more oversight. This AI overhaul is expected to power a suite of practical new features. These include AI-powered photo editing to remove objects, automatic summarization of notifications and web pages, and the transcription of voice memos. The move to partner with Google follows reports that Apple, while a leader in on-device processing, needed to accelerate its capabilities in the large-scale generative AI race.
