Apple is reportedly integrating Google's Gemini extensively into its new Siri, while running all computations on Apple's private cloud compute (PCC) servers to ensure data privacy and security.
Bloomberg cited insiders saying Apple has finalized the technical architecture for the updated Siri. The company will pay Google to develop a customized version of Gemini that can securely operate within Apple's PCC environment, aiming for a spring 2026 launch alongside iOS 26.4.
Rather than directly adopting Google services, Gemini will power Siri's intelligent responses behind the scenes under Apple's own AI branding, with no visible Google logos or references to users.
Earlier rumors indicated the Siri overhaul involves three main components: query planning, knowledge search systems, and summary generation. Gemini will handle core computations for the first two, enhancing Siri's deeper understanding and semantic organization capabilities.
Tech outlets like 9to5Mac and Computerworld note this approach mirrors Samsung Electronics' strategy of using Gemini to support AI features in Galaxy phones, a pragmatic move not strictly tied to fully in-house technology development.
This arrangement allows Apple to rapidly boost AI capabilities without compromising its strict privacy principles, filling gaps where its proprietary models are still maturing and meeting user expectations for a smarter Siri. Meanwhile, Google expands Gemini's market share, creating a win-win scenario.
Apple continues developing its own AI systems but has accelerated product rollout by partnering with multiple AI firms, including integrating OpenAI's ChatGPT into Apple Intelligence and enabling developers to use Anthropic Claude models in Xcode 26. In China, Apple collaborates with Baidu and Alibaba as well.
Article edited by Jack Wu
