At WWDC Apple unveiled its most substantial Siri overhaul in years, branding the suite as Apple Intelligence. The upgrade leverages techniques from Google Gemini, a collaboration announced earlier this year. New capabilities include richer voice commands, natural‑language Shortcut creation, and more accurate dictation across iOS 27 devices. Apple emphasizes that every request runs on‑device or through its encrypted private cloud, keeping user data private.

Apple says four proprietary Foundation Models drive the service: AFM Core, comparable to Gemini Nano for on‑device tasks; AFM Core Advanced adds multimodal support; and two cloud‑based models handle heavier image and language requests. All processing runs on‑device when possible, otherwise through Private Cloud Compute, ensuring user data never leaves the device or is shared with Google.

The partnership lets Apple accelerate AI rollout without building a large language model from scratch, but it also ties iOS 27’s reputation to Gemini’s performance. Consumers gain faster, more context‑aware interactions, while privacy claims hinge on Apple’s private cloud architecture. Real‑world usability will determine whether the upgrade justifies the hype. Analysts view the collaboration as evidence that major players will share foundational AI research rather than battle on every layer.