Apple Rebuilds Siri, Adds System-wide Search Gesture

Per Bloomberg reporter Mark Gurman, iOS 27 will ship with a "completely rebuilt" Siri that surfaces a new system-wide search gesture and a hybrid voice/chat interface. Gurman reports users can activate a system search by swiping down from the top center of the screen; that action will surface a "Search or Ask" bar in the Dynamic Island with a microphone icon to enable voice input. Gurman writes the search UI will allow users to "press the search bar to toggle between using Siri or third-party offerings like ChatGPT or Gemini as their search engine."

Gurman also reports the interface will permit uploads of images and documents via a "+" button and that users can swipe a transparent results card down to enter a chatbot-like conversation mode with in-line mini app cards for things like weather, appointments, and notes. These details are reported in Bloomberg and summarized by outlets including 9to5Mac.

The reporting frames this Siri update around an agentic, multimodal assistant and a system-level search affordance. Separately, Google Cloud CEO Thomas Kurian said at Google Cloud Next '26 that Google is "collaborating with Apple as their preferred cloud provider to develop the next generation of Apple Foundation Models based on Gemini technology," a quote reported by T3 and India Today. That attribution links the reported Siri redesign to Google's Gemini family as the underlying foundation for Apple's new models.

Editorial analysis: For practitioners, two patterns matter. First, moving an assistant to a persistent, system-level gesture plus a chat-thread UI mirrors trends in mobile OS design where assistants become primary interfaces for information retrieval and task execution. Second, the reported vendor model arrangement with Google Cloud and Gemini highlights a separation between device OS UI changes and cloud model sourcing, with implications for latency, on-device vs cloud inference, and data flow.

Editorial analysis: Reported support for toggling third-party search backends and uploading images/documents raises integration and privacy questions commonly encountered when platforms open system services to external models. Companies that enable third-party LLMs for system search typically need standardised APIs, permissioning, and logging practices. Observers will want to know whether Apple will publish SDKs or APIs to let developers integrate model-driven search results into the new UI; the reporting to date does not cite an Apple developer announcement.

Editorial analysis: Key indicators to follow ahead of WWDC are: whether Apple publishes developer guidance for the new search gesture and chat UI; whether Apple documents on-device vs cloud inference responsibilities for Gemini-backed features; privacy and data residency language tied to third-party backend toggles; and any SDKs or entitlements for third-party LLMs. Also watch for Bloomberg's full reporting at WWDC and official Apple developer notes, which will confirm implementation and any limitations.

The specific implementation details above come from Bloomberg reporting summarised by outlets including 9to5Mac. Apple has not provided an official developer release or detailed public documentation of these features in the sources cited here.