Apple unveils “Siri AI,” rebuilding its voice assistant around Apple Intelligence

Apple on Monday moved more decisively into the generative AI race by announcing “Siri AI,” an entirely new version of Siri powered by its next generation “Apple Intelligence” platform. Apple framed the upgrade as more than a voice command tool, positioning Siri AI as a conversational assistant that can understand personal context, see what is on a user’s screen, and use both user data and web information to answer questions and help complete tasks across apps. The company said Siri AI is available for developer testing starting immediately, with a beta for users later this year.

In Apple’s description, Siri AI is designed to handle “back-and-forth” questions, pulling from sources such as messages, emails, photos and other stored content to find relevant details when users ask. The assistant can also address questions tied to content displayed on the screen, and it can go out to the web for up-to-date information using what Apple calls “broad world knowledge.” Apple also announced a dedicated Siri app for users to revisit past conversations across devices, and it added Visual Intelligence and integrated writing tools designed to generate and revise text directly in places where people type.

Apple said the Siri AI experience is built on a “bold new architecture” for Apple Intelligence that is designed around privacy protections. The press materials say Siri AI leverages Apple Foundation Models that run both on device and on servers using “Private Cloud Compute.” Apple also said that when Private Cloud Compute is handling requests, users’ personal data is not stored or made accessible to Apple or others, and that outside experts can verify the privacy promise. Under the hood, Apple described system tools such as the Spotlight index and an “App Toolbox” that operate on device as part of how Siri AI orchestrates responses.

Apple emphasized how the assistant will appear and work across its ecosystem. On iPhone, the company said users can start a conversation by swiping down from the Dynamic Island, or by using other standard activation methods like the side button. On iPad and Mac, Siri AI is integrated into Spotlight search, and Apple said it can be invoked through systemwide context menus so users can control-click images, files, or text. For Apple Vision Pro, Apple described a spatial interface that lets users place a 3D visualization in their environment and invoke Siri AI by looking at it and speaking.

Beyond Siri AI itself, Apple’s “Apple Intelligence brings powerful AI capabilities into everyday experiences” announcement outlined a wider set of features intended to make common apps smarter. Apple said the next generation of Apple Intelligence brings deeper editing capabilities in Photos, new tools for browsing in Safari, and an all-new Image Playground that can generate high-quality images, including photorealistic output delivered via Private Cloud Compute. Apple also said generated images include a hidden SynthID watermark intended to identify them as AI-generated, and it described systemwide improvements such as intelligent proofreading and suggestions for naming files and folders.

Apple also linked the AI announcements to other software changes rolling through the same release cycle. In a separate WWDC preview release, Apple said it would add parental controls and significant updates to Screen Time, plus performance and design improvements across its platforms, including faster app launches and faster photo loading. However, Apple did not give a single, specific “consumer launch day” for Siri AI in that preview, instead describing a developer testing start and a later beta timeframe, and it provided regional limitations that could affect how broadly the features roll out.

What to watch next is how quickly Apple translates the announcements into working, integrated experiences for developers and users. Apple said Siri AI features will be available for developer testing starting June 8, 2026, and it will be available as a beta to users later this year on supported devices and in supported languages. Apple also said some regions have restrictions, including an initial lack of Siri AI availability in the EU on iPhones and iPads under iOS and iPadOS, while Reuters reported that Siri AI will not be available in China at launch as Apple works through regulatory issues. Apple’s bet is that tighter OS integration, privacy-focused deployment, and on-screen and personal context awareness will be enough to make Siri AI competitive as rivals push “agentic” AI into everyday computing.