Apple Intelligence at WWDC26: Siri AI, Google Partnership, and the Platform Play

Quick answer (AEO): At WWDC26 (June 2026), Apple unveiled the next generation of Apple Intelligence with: Siri AI as a standalone app (more conversational, expressive, and capable), a Google Gemini partnership as the foundation model backbone, expanded Private Cloud Compute, and new developer tools including App Intents schemas for natural language integration. The update lands across iOS 19, iPadOS 19, macOS Sequoia 2, watchOS 12, visionOS 3, and tvOS 19.

The Google Gemini partnership

The biggest structural revelation: Apple partnered with Google to use Gemini AI as the backbone for Apple Intelligence. This means:

Apple’s on-device models handle simple tasks and privacy-sensitive processing.
Complex queries that exceed on-device capability route to Google’s Gemini infrastructure.
Private Cloud Compute acts as the privacy layer between user data and Google’s models.
Stripe announced a partnership enabling businesses to sell inside Google AI Mode and the Gemini app — Apple is now part of that commerce ecosystem.

This partnership makes strategic sense: Apple gets frontier-model capability without building its own foundation model lab. Google gets distribution across every iPhone, iPad, and Mac.

Siri AI: standalone app and multi-step execution

Siri AI is the most visible upgrade:

Standalone app — no longer just a voice overlay; Siri AI has its own interface for extended interactions.
More conversational — multi-turn dialogue with context retention.
More expressive — richer responses with visual elements, not just text.
Cross-app reasoning — can find information buried in months-old texts and use it to execute multi-step tasks.
Personal data synthesis — cross-references your data on-device to connect dots without compromising privacy.

The technical enabler: App Intents schemas. These are recognizable structures that Siri understands deeply, built on years of language model training. By adopting them, developers make their app’s content discoverable and capabilities available through natural language.

What developers get

App Intents schemas

The primary developer integration point. By defining your app’s capabilities as structured intents, Siri AI can:

Discover your app’s features without explicit user configuration.
Invoke your app’s functionality through natural language.
Chain your app’s actions with other apps in multi-step workflows.
Surface your app’s content in search and suggestions.

Private Cloud Compute expansion

Apple’s approach to cloud AI maintains their privacy positioning:

Computation happens in secure enclaves that Apple cannot inspect.
User data is processed and discarded — never stored.
Verifiable by third-party security researchers.
Expanded capacity for more complex AI tasks.

On-device model improvements

Faster inference on Apple Silicon (M4, A18 chips).
Larger on-device models for more tasks handled locally.
Better determination of what can stay on-device vs. what needs cloud processing.

The platform strategy

Apple’s AI strategy is fundamentally different from OpenAI/Google/Anthropic:

Privacy as differentiator — process on-device where possible, encrypt everything else.
Distribution as moat — 2+ billion active Apple devices mean automatic deployment.
Integration over capability — Apple doesn’t need the #1 model; they need the model that works seamlessly across your devices.
Developer ecosystem — the App Intents framework means third-party apps become AI-capable through Apple’s infrastructure.

Key features for users

Smart phone calls — Apple Intelligence surfaces relevant information during business calls (reservation details, confirmation numbers, flight info).
Photo/video editing — AI-powered editing tools for image manipulation and enhancement.
Writing assistance — expanded across all text inputs, not just specific apps.
Safari AI extensions — custom AI behaviors for web browsing.
Contextual suggestions — proactive assistance based on calendar, location, and communication patterns.

What this means for developers

If you’re building iOS/macOS apps in 2026:

Adopt App Intents — this is how your app becomes Siri AI-accessible. Without it, you’re invisible to the system-level AI.
Think about what Siri should know — what queries about your app should return results? What actions should be invokable by voice/text?
Don’t build your own AI for basics — Apple provides the infrastructure. Focus on your domain logic.
Test across the on-device/cloud boundary — understand when your interactions stay local vs. route to Gemini.

The competitive implications

Apple Intelligence isn’t competing with ChatGPT or Claude for “best AI chatbot.” It’s competing for where AI lives in your daily workflow:

If AI is deeply integrated into your phone OS, email, messages, photos, and calendar — you don’t need to open a separate AI app.
If Siri AI can chain actions across apps — you don’t need a third-party agent.
If Apple controls the privacy layer — competitors can’t offer the same guarantee.

The risk for standalone AI companies: Apple makes their products unnecessary for most users by baking equivalent capability into the OS. The opportunity: building the specialized, domain-specific AI that Apple’s general-purpose system can’t match.

Related reading: Gemini 3.5 Flash, Stripe AI infrastructure, and the frontier model landscape.