Private Cloud Compute

Siri AI, Apple's overhauled assistant rolling out in iOS 27, is generating user complaints over network errors, broken device search, unlock requirements, slow Apple Watch responses, and missing features from classic Siri. The assistant remains labeled beta with no disclosed fix roadmap beyond new language support in October.

Apple released iOS 27 on Monday, introducing a beta version of its redesigned Siri AI assistant ahead of iPhone 18 availability this week. The new Siri uses large language models to search personal data like texts and calendars, but requires opt-in and isn't yet available in the EU or China.

Apple has confirmed that its new Siri AI features in China will be powered by Alibaba's Qwen models rather than Google Gemini, which handles Siri AI elsewhere. A published guide indicates users will need separate Qwen accounts, suggesting the privacy protections built into Apple's Private Cloud Compute won't apply.

Apple announced five third-generation foundation models at WWDC26, headlined by AFM 3 Core Advanced—a 20-billion-parameter sparse model that runs on-device by activating only 1-4 billion parameters at a time. For the first time, Apple extended Private Cloud Compute to third-party infrastructure, with AFM 3 Cloud Pro running on Nvidia GPUs in Google Cloud.

Apple executives revealed at WWDC that its most advanced AI model, AFM Cloud Pro, runs on Nvidia GPUs deployed in Google's cloud infrastructure while maintaining Apple's privacy guarantees. The company disclosed it uses Google's Gemini frontier models to refine its own custom models built for Apple Silicon.

Apple's senior vice president Craig Federighi disclosed that iOS 27's Siri AI uses a family of third-generation Apple Foundation Models trained with outputs from Google's Gemini frontier models. The most capable model, AFM Cloud Pro, runs on Nvidia GPUs in Google's cloud infrastructure while maintaining Apple's Private Cloud Compute privacy architecture.