Apple introduced a conversational overhaul for its voice assistant called Siri AI during its WWDC keynote. Rolling out this fall, the system leverages personal context to search through emails and messages for specific data. This allows the assistant to handle multi-step tasks, such as planning a party menu based on a friend's recommendation found in a chat.

Technical requirements create a divide in user experience. The most capable model requires high-end hardware, specifically M3 chips or later on Macs and M4 CPUs on iPads with at least 12GB of memory. Users with older devices receive a stripped-down version that lacks certain capabilities, including the more expressive, customizable voice options and improved dictation accuracy.

Integration extends across the ecosystem via a new Siri app and Visual Intelligence for camera-based queries. While some tasks run locally, others rely on private cloud compute, meaning image generation will face daily usage limits. Google provides the underlying Gemini model for parts of this update, though Apple maintains that privacy protections remain non-negotiable for all users.

These updates bring Siri closer to a true personal assistant by linking on-screen awareness with deep app integration. The system can now extract calendar dates from images or rewrite text to match a user's specific writing style.