Apple finally shipped the Siri AI overhaul. Two years after promising a smarter Siri at WWDC 2024 and missing every deadline since (I tracked the delays here), the company showed the real thing on stage at Tim Cook’s final keynote on Monday.
It’s called Siri AI. It’s a complete conversational AI assistant that goes head to head with ChatGPT, Claude, and Gemini. It has a dedicated app. It lives in the Dynamic Island. It can see what’s on your screen, pull information from your messages and emails, and write documents in your voice.
And it’s not coming to Europe.
Here’s the full rundown.
What Siri AI Actually Does
Siri AI is a complete rewrite. Apple rebuilt Siri from the ground up on the next generation of Apple Intelligence, which, as reported in January, runs on Google’s Gemini models under the hood.
The key capabilities:
- Conversational AI. Full back-and-forth dialogue. You can brainstorm, ask follow-up questions, get in-depth plans. It’s not a one-shot voice command anymore.
- Personal context understanding. Siri can pull information from your messages, emails, photos, calendar, and contacts. Ask it to find that restaurant recommendation a friend texted you, surface a hotel confirmation from an old email, or pull up photos from your last trip.
- Onscreen awareness. Siri can see what’s on your screen and answer questions about it. Reading a text about a potluck? Ask Siri what to bring and it’ll add a recipe to Notes.
- Broad world knowledge. Real-time web grounding for up-to-date answers. When’s the next solar eclipse? What’s the setlist for that tour?
- Write with Siri. Compose emails, messages, and documents by describing what you need. In Mail and Messages, it adapts to how you write to each person. If you send your boss short bullet points, that’s what it drafts.
- Visual Intelligence. Point your camera at something and Siri identifies it, splits bills, gives nutritional info. Coming to iPad and Mac for the first time too.
It Has a Dedicated App
There’s a new Siri app where all your conversations live. It syncs across devices via iCloud, so you can start a conversation on your Mac and continue it on your iPhone. That alone signals how serious Apple is about this being a platform, not just a voice feature.
The Dynamic Island Gets a Job
Siri AI lives in the Dynamic Island on modern iPhones. Swipe down from the Island to start typing a conversation, use the side button, or say “Hey Siri” like always. The old screen-glow animation is replaced by a new Dynamic Island animation.
On Mac, Siri is baked into Spotlight. On Apple Watch, you can start a conversation from your wrist. On Vision Pro, there’s a spatial 3D visualisation you look at to invoke Siri.
The Architecture and the Gemini Question
According to Apple’s official press release, Siri AI runs on-device Apple Foundation Models where possible, and Private Cloud Compute for requests that need more power. Apple says your data is not stored or accessible to Apple on PCC.
But the elephant in the room: Siri AI runs on Google Gemini. Apple confirmed the partnership in January, and the Siri AI press release dances around it, talking about “the next generation of Apple Intelligence” and “the system orchestrator” without naming the model provider directly. The architecture diagram shows on-device models connecting to a “system orchestrator” that taps into Spotlight and App Toolbox. The heavy lifting comes from Google’s servers.
If you’ve been following Google’s personal intelligence rollout, the overlap is obvious. Apple bought what Google built.
When You Can Use It
| Platform | Availability |
|---|---|
| Developer beta | Today (June 8), iOS 27, iPadOS 27, macOS 27, visionOS 27 |
| Public beta | “Later this year” |
| iPhone devices | iPhone 16+, iPhone 15 Pro/Max |
| iPad | iPad mini (A17 Pro), iPad M1+ |
| Mac | Apple Silicon M1+ |
| Other | Apple Watch Series 10+, Ultra 2+, SE 3; Apple Vision Pro |
| Language | English only at launch (more coming) |
| EU (iOS/iPadOS) | Not available due to DMA regulatory issues |
| EU (Mac/Watch/Vision) | Available |
| China | Not available |
The EU situation is the biggest asterisk. Apple published a separate press release titled “Due to DMA, Siri AI delayed in EU for iOS 27 and iPadOS 27.” Mac, Apple Watch, and Vision Pro users in the EU can still access it. Apple says it’s “working hard to find a path forward that preserves users’ privacy and security.”
China is also excluded while Apple works through “regulatory requirements.”
Tim Cook’s Final WWDC
This keynote doubled as a farewell. Tim Cook presided over his 15th and final WWDC. Incoming CEO John Ternus takes the reins in September. Cook closed with: “It’s been the honour of a lifetime. I truly believe the best is still ahead.”
The Siri AI launch is the last major product announcement of the Cook era. Whether the best is still ahead depends on whether this actually ships to users this time, whether the Gemini dependency becomes a liability, and whether Apple can resolve its EU standoff before the competition widens the gap further.
The Bottom Line
I wrote back in December that Apple needed to commit to Siri or stop selling the dream. Today, they committed.
Siri AI is real. It’s a genuine ChatGPT competitor. It has the deep system integration only Apple can pull off: Dynamic Island, Spotlight, systemwide dictation, watchOS, Vision Pro. But it’s two years late, it runs on Google’s models, and large parts of the world won’t get it at launch.
That’s the Apple AI story in 2026. They finally shipped. The caveats are the story.
I’ll update this piece when the public beta drops with first-hand impressions.
