Apple Is Building an AI App Store Inside Siri. Most Developers Haven't Noticed Yet.

Remember the early days of the App Store? 2008. The gold rush. Some guy made a flashlight app and bought a car with the proceeds. Another one sold a virtual lighter for $0.99 and made $200,000 in a month. The bar was low, the demand was astronomical, and the people who showed up first got rewarded disproportionately.

We’re about to see that happen again. Except this time, the store isn’t for apps. It’s for AI.

And it’s hiding inside Siri.

Siri Is Becoming a Chatbot. With a Storefront.

Let’s start with what’s actually happening, because the headlines have been all over the place.

iOS 27 — expected to be unveiled at WWDC on June 8 — is bringing a completely rebuilt Siri. Not “Siri with better responses.” Not “Siri but it understands follow-up questions now.” A full ChatGPT-style chatbot app that lives on your Home Screen.

Internally codenamed “Campos,” this standalone Siri app will look like Apple’s Messages app, but for conversations with AI. You’ll see a list of past conversations. You can favorite chats, search through them, start new ones, upload images and documents. It’ll support both voice and text. The conversations will even auto-delete after 30 days if you want — a privacy feature that’s very Apple.

But here’s the part that matters for developers: Apple is building an Extensions system that lets third-party AI chatbots plug directly into Siri. Claude, Gemini, ChatGPT — they’ll all be available as alternatives. Users will be able to route different types of queries to different providers. Gemini for research. Claude for coding. ChatGPT for creative writing.

And Apple is creating a dedicated Extensions section inside the App Store to showcase these integrations.

Read that last sentence again. A new section. In the App Store. For AI integrations. That’s not a feature update. That’s a distribution channel.

Why This Is Bigger Than Widgets, App Clips, or Live Activities

Every few years, Apple opens a new surface area for apps. Widgets in 2020. App Clips in 2020 (remember those? No? Exactly). Live Activities in 2022. Each one creates a brief window where early adopters get outsized visibility.

Siri Extensions is different because of scale.

Widgets live on the Home Screen — a place users already curate. App Clips require a physical trigger (NFC tags, QR codes). Live Activities compete for a tiny strip of screen real estate.

But Siri? Siri is the default AI interface on 2 billion Apple devices. Every iPhone, iPad, Mac, Apple Watch, HomePod, and Vision Pro. When Apple ships this chatbot interface, it becomes the first thing billions of people try when they want to talk to AI. And when those people decide they want something more specialized than the default, they’ll browse the Extensions section.

This is like the early App Store, except the user base is 50x larger from day one.

Think of it this way: if the original App Store was a new shopping mall in a growing suburb, the Siri Extensions marketplace is a pop-up shop inside the busiest airport in the world. Everyone’s already there. They’re already looking. You just need to have something on the shelf.

What Should You Actually Build?

OK, so there’s a new marketplace coming. Cool. But what goes in it?

We don’t have the full Extensions API documentation yet — that’ll come with the WWDC sessions. But based on what we know about Apple’s approach (and the Foundation Models framework that shipped with iOS 26), here’s where the opportunities are:

Vertical AI assistants. The big players — Claude, Gemini, ChatGPT — are general-purpose. They’re good at everything, great at nothing specific. The real opportunity is in specialized AI that does one thing exceptionally well for a specific audience. A cooking assistant that plans meals and generates shopping lists. A fitness coach that adapts workouts based on Apple Health data. A study buddy for students — something like what we built with ThinkBud, our AI-powered learning companion that helps you break down complex topics using the Feynman technique.

Prompt management layers. Most people are terrible at talking to AI. They don’t know how to write good prompts. They don’t know how to iterate. A Siri Extension that acts as a prompt intermediary — translating vague requests into structured, effective queries — has massive potential. This is something we’ve been exploring with PromptKit, and the Siri Extensions system could be the perfect distribution vehicle for it.

Domain-specific knowledge bases. Imagine a legal AI that understands your jurisdiction. A medical AI trained on patient communication guidelines. A real estate AI that knows local market trends. These are the kinds of extensions that justify a subscription and build genuine lock-in.

Workflow integrations. The most powerful version of this isn’t “another chatbot” — it’s an AI that connects to the things you already use. Your calendar, your notes, your project management tool. Apple’s already shipping on-device intelligence through the Foundation Models framework. Extensions that bridge that local context with specialized cloud AI will be the killer apps of this platform.

The Architecture Problem (And Why You Should Start Learning Now)

Here’s where most indie developers are going to trip: building a Siri Extension isn’t just about wrapping an API call. It requires understanding App Intents, which is Apple’s framework for making your app’s functionality discoverable to Siri and Shortcuts.

If you haven’t touched App Intents yet, you’re behind. Not critically — you’ve got until the iOS 27 beta this summer — but the learning curve is real.

You’ll need to:

Define your app’s capabilities as structured intents. Siri needs to know what your app can do, what parameters each action takes, and what results it returns.
Handle context from Siri conversations. Users won’t be launching your app directly — they’ll be mid-conversation with Siri when your Extension gets invoked. Your code needs to handle that gracefully.
Integrate with the Foundation Models framework for any on-device processing. Free inference, no API keys, no cloud costs. If you’re not using this yet, you’re leaving money (and user trust) on the table.
Build with privacy as a constraint, not an afterthought. Apple’s already announced that Siri conversations will run through private cloud compute. Your Extension will be held to the same standard.

If you’re looking to get up to speed on the architectural foundations — modular Swift packages, App Intents integration, composition patterns that scale — our SwiftUI at Scale course covers exactly this kind of architecture. The first few lessons on SPM feature modules and public API design are directly relevant to how you’d structure an Extension.

We wrote about Apple opening Siri to third-party AI back in April, and the Core AI framework changes coming at WWDC. But the Extensions marketplace angle is the one I keep coming back to. It’s the business opportunity hidden inside the technical announcement.

The Window Is Small. The Payoff Is Massive.

Here’s the uncomfortable truth: this window won’t last long.

When the App Store launched, it took maybe 18 months before the market got saturated. Before every niche had five competitors and the only way to get noticed was spending $50,000 on Apple Search Ads.

The Siri Extensions marketplace will compress that timeline even further. The big AI companies will move fast. The VC-funded startups will throw money at it. Within a year of launch, the good real estate will be claimed.

But right now? Right now, most developers are focused on WWDC’s flashier announcements — the foldable iPhone APIs, the Core AI framework, the Siri redesign itself. The Extensions marketplace is getting treated as a footnote. A “we’ll figure it out when the API docs drop” afterthought.

That’s your advantage. The developers who are studying App Intents today, building with Foundation Models today, designing their app’s capabilities as structured intents today — they’ll be the ones shipping on day one of the iOS 27 beta. While everyone else is still reading the documentation.

The flashlight app era is over. But the AI extension era is just starting.

The Takeaway

Apple is about to do something it hasn’t done since 2008: open an entirely new category of the App Store. The Siri Extensions marketplace is a fresh distribution surface on 2 billion devices, and the developers who treat it as a platform opportunity — not just a Siri feature — will be the ones who benefit most.

WWDC is 17 days away. The API docs are coming. The question isn’t whether this matters. It’s whether you’ll be ready when it does.

If you’re building iOS apps and haven’t started thinking about how your app’s intelligence could surface through Siri, now’s the time. Dust off App Intents. Experiment with the Foundation Models framework. And start imagining what your app looks like when Siri is the front door.

The best storefronts get the best foot traffic. And this storefront is about to get very, very crowded.