Apple and Google AI Deal: Google Gemini Distillation for On-Device AI

Apple has access to Google's Gemini AI model, allowing it to customize the model for Siri and other AI features. Apple can distill the massive model into smaller, more efficient chunks for on-device AI processing, enabling the iPhone to run AI tasks offline.

The details of the AAPL-GOOGL AI partnership are deeper than previously understood, according to a report from The Information . Apple has full access to Google's Gemini AI model within its own data centers, granting it the ability to customize and adapt the model for specific on-device applications, including a major overhaul of Siri expected to debut at WWDC this June.

Central to the arrangement is Apple's ability to "distill" Gemini, a process in which knowledge from the large foundation model is transferred to smaller, more efficient models capable of running directly on iPhones and other Apple devices . These distilled models require less computing power and deliver faster results, enabling offline AI capabilities that reduce Apple's dependence on cloud-based processing. The approach represents a pragmatic strategy for Apple, which has lagged competitors in deploying advanced AI features.

The deal carries significant strategic implications for both companies. For Apple, it provides a shortcut to competitive AI capabilities ahead of WWDC 2026, while Google secures a massive distribution channel for its Gemini technology across billions of Apple devices. However, Apple's growing reliance on Google's AI infrastructure raises questions about long-term independence and the competitive dynamics between the two tech giants.