AI Models & Platforms
Apple Opens Siri AI Developer Preview for Custom App Integrations

Apple announced on November 3, 2026 that its Siri AI Developer Preview is open to all developers through the iOS 27.3, iPadOS 27.3, macOS 27.3, and visionOS 27.3 betas. Apps can expose custom content and actions through App Intents, allowing Siri AI to handle requests such as checking deliveries, booking reservations, and placing orders.
The opening extends Siri AI beyond actions and data that fit Apple’s system-defined app schemas. Integrations can draw on personal-context understanding and onscreen awareness, letting people phrase requests in their own words, and Siri AI can surface any entity in Spotlight’s semantic index rather than only entities matching an existing schema. Apple has begun rolling out related capabilities across its own apps, including Calendar, Files, Mail, Maps, Messages, Photos, and Shortcuts; examples include finding a previously visited coffee shop and archiving files unopened for more than a year.
How Apps Appear in Siri AI
Under the Developer Preview documentation, invocation depends on context. A person does not need to name an app while it is in the foreground, while a Live Activity is active, or when continuing a Siri conversation that already included the app. Background requests require the app’s name. For example, a ride-hailing customer could follow a Live Activity and then ask Siri to change the pickup address without restating the app.
Siri AI can also find content in one app, transfer it to a specified second app, and perform an action there. Apple’s example sends a user’s favorite sneakers through a messaging app after resolving the recipient through Contacts. Requested custom content can appear in developer-supplied card snippets that use the app’s own views and branding. No new entitlement or separate developer program is required.
Developer Controls and App Skills
Apple’s implementation guidance makes exposure selective: developers decide which content Siri may retrieve and which actions it may perform. A Swift macro, @AppIntent(.tool), turns a new or existing custom intent into a tool Siri can call. Intents already conforming to a Siri-supported schema in a primary domain do not need that annotation.
Titles and descriptions for intents, parameters, and entity properties help Siri interpret natural-language wording, distinguish similar actions, and formulate responses. Developers can index entities in Spotlight, associate visible app content with entities for onscreen context, and donate intents when users perform matching actions to supply behavioral cues. Transferable representations allow custom entities to move between apps during cross-app requests.
Access can be narrowed for sensitive data. Authentication policies can protect an entire entity or selected properties, while an internal visibility setting prevents Siri, Shortcuts, and other system features from reading a property. Apple’s finance example requires authentication for an account nickname and balance but keeps the routing number internal to the app and its extensions.
For multi-step work, app skills supplement individual tool intents. A skill is a Markdown file whose metadata helps Siri decide whether it matches a request; Siri consults the instruction body only after making that match. The instructions can describe action order, which intent to prefer, practices to avoid, app-specific domain knowledge, and when to ask clarifying questions. A shopping skill could coordinate searching for items, adding them to a cart, choosing payment and delivery methods, reviewing the cart, and confirming the purchase.
Testing and Availability Limits
Xcode’s Siri Integration Advisor offers implementation feedback, and the Siri Evaluations framework lets developers script requests, measure responses, and detect changes as intents and entities evolve. Developers can run development-signed builds on supported devices and share them with internal TestFlight testers. External TestFlight testers cannot use the preview, and its functionality does not work in App Store versions even when those apps contain the new integration.
Apple says the first customer release is expected in a few months, with App Store submissions scheduled to open in the spring. Further terms and usage policies for Siri AI-enabled experiences will follow. Siri AI requires an Apple Intelligence-enabled device set to a supported language. It is available in English, French, Japanese, Korean, Portuguese, and Spanish, with Traditional Chinese, Norwegian, Turkish, and Vietnamese listed as forthcoming.
Siri AI is unavailable in the EU on iOS, iPadOS, and watchOS, and Apple says it is working with the EU on a possible route to availability. It is not available to users under 13, and some features may not be available in every region or language.
Certain Apple Intelligence features that rely on server-side models, including Siri AI, are subject to daily usage limits. Apple says expanded access to those features will be available for a fee in the future.
The immediate change is a build-and-test opening rather than a full customer release. Developers can define the data and actions Siri AI may use, orchestrate them across workflows, and evaluate the results internally, while customer availability, App Store submissions, and detailed usage policies remain tied to later stages of Apple’s rollout.












