Apple May Use Google Cloud for AI Siri Upgrade
Apple is reportedly considering using Google's cloud infrastructure to host data for its upcoming AI-powered Siri assistant. The potential move underscores the massive server capacity needed for next-gen AI and suggests that even tech giants are forming surprising alliances to compete.
The multi-year deal officially makes Google's Gemini models a core component of Apple Intelligence, aiming to supercharge Siri's capabilities. This partnership is geared towards providing a more personalized and contextually aware Siri by late 2026. The agreement is seen as a way for Apple to accelerate its AI development and reduce the risk of falling behind in the competitive AI landscape. Under the arrangement, Apple will leverage Google's advanced cloud technology but will run the Gemini models on its own Private Cloud Compute infrastructure. This is designed to ensure that Google does not have direct access to user data, a key component of Apple's privacy-focused strategy. For more complex requests, Siri will have the option to tap into the capabilities of partners like Google, but this will be an opt-in feature for users. This collaboration extends a long-standing and lucrative, if sometimes contentious, relationship between the two tech giants. For years, Google has paid Apple billions annually—an estimated $20 billion per year recently—to be the default search engine in the Safari browser. This existing financial arrangement has faced antitrust scrutiny. While embracing Google's large-scale AI, Apple continues to develop its own on-device artificial intelligence under the codename "Project Greymatter." These on-device models are designed to handle tasks like summarizing notifications and will work in tandem with the more powerful cloud-based AI from its partners. The move to use external cloud infrastructure marks a significant strategic shift for Apple, which has historically prioritized building and controlling its own server farms. This hybrid approach, blending its own private cloud with resources from rivals like Google and Amazon Web Services, underscores the immense infrastructural demands of cutting-edge AI. Analysts suggest the deal is a win-win: Google gets its Gemini models on over two billion active Apple devices, a massive distribution victory, while Apple gets to rapidly deploy advanced AI without the massive upfront capital expenditure of building comparable systems from scratch. The licensing agreement is estimated to cost Apple around $1 billion annually.
