WWDC 2026 Is 26 Days Away. This Is Your iOS Developer Pre-Flight Checklist.
You know that scene in The Martian where Matt Damon realizes he’s alone on Mars, takes one breath, and then opens a checklist? He doesn’t panic. He doesn’t write a blog post about how NASA abandoned him. He starts ticking boxes. Potatoes. Water reclamation. Duct tape everything.
That’s you right now. WWDC 2026 is 26 days away. The keynote lands June 8. And if you haven’t looked at the leaks, the reports, and the increasingly non-subtle hints Apple has been dropping, your codebase might be sitting on a pile of deprecation warnings that are about to become deprecation obituaries.
I went through every credible report, every leaked detail, every suspiciously specific Gurman sentence. Here’s the pre-flight checklist — five things you should be doing this week, not during the keynote livestream while pretending to take notes.
1. SiriKit is getting its retirement party. Move to App Intents.
This is the big one, and if you’ve been reading this blog, you saw it coming. We called it two weeks ago: App Intents are the new SEO, and SiriKit is the old phone book nobody opens anymore.
Multiple reports now point to a formal SiriKit deprecation notice at WWDC 2026. Apple typically gives you 2-3 years before deprecated frameworks actually stop working, so it’s not an emergency — it’s a deadline with a very generous countdown timer that most developers will ignore until year three.
Don’t be that developer.
What to do this week: Open your project, search for import Intents. If you find it, you have work to do. The migration to AppIntent protocol isn’t a rewrite — it’s a restructure. Your intent definitions get shorter, your parameter handling gets cleaner, and suddenly Siri can actually find your app. We wrote a complete walkthrough with real code if you want the step-by-step.
2. Core ML is becoming Core AI. Isolate your ML code now.
Apple’s Core ML framework shipped in 2017. It was built for a world where “AI” meant image classification and sentiment analysis. We’re not in that world anymore, and Apple knows it.
Reports from Bloomberg’s Mark Gurman and 9to5Mac point to a Core AI framework in iOS 27 — built for LLMs and generative AI, supporting both on-device and cloud-based models, with the system deciding where to run inference based on performance and privacy.
We covered this in detail back in April, and the advice hasn’t changed: Core ML apps won’t break overnight, but the migration is coming.
What to do this week: If you use Core ML, put a protocol in front of it. Something like AIServiceProtocol with predict(input:) and generate(prompt:). When Core AI drops, you swap the implementation behind the protocol. Your views, your view models, your tests — none of them care. This is the kind of boring architecture work that pays for itself in a single WWDC week.
If you’re not using Core ML yet but want to start building with Apple’s Foundation Models, wait three weeks. Core AI will likely give you a cleaner entry point than what’s available today.
3. Liquid Glass isn’t going anywhere — but it IS getting fixed.
Apple dropped the Liquid Glass design language with iOS 26 and Tahoe, and the reaction was… mixed. Translucency that made text hard to read. Sidebars that looked like they were melting. Forums full of accessibility complaints.
Good news: Apple listened. The Apple Developer app just got a Liquid Glass redesign that looks significantly cleaner — better contrast, sharper shadows, more readable text over translucent surfaces. That redesign is a preview of where macOS 27 and iOS 27 are headed.
Bad news: if your app has custom UI components that fight Liquid Glass, they’re about to look even more out of place.
What to do this week: Run your app on iOS 26. Look at every custom toolbar, every navigation bar override, every place where you set a background color because the default “didn’t look right.” Those hacks will probably clash with the iOS 27 refinements. Replace solid backgrounds with .ultraThinMaterial or .regularMaterial where possible. Lean into the system’s design, don’t fight it. We explored three approaches to Liquid Glass app icons that show how to work with the aesthetic instead of against it.
4. Foldable iPhone APIs are real. Audit your layout assumptions.
This one sounds like future-proofing until you realize Apple is doing something unprecedented: shipping developer APIs before the hardware exists. We wrote about this in April and the reports keep getting more specific. Expect new SwiftUI and UIKit APIs for adaptive layouts, hinge state detection, and multi-configuration screens.
Even if you don’t care about the foldable iPhone, these APIs signal where Apple is pushing layout design: flexible, adaptive, screen-agnostic.
What to do this week: Run your app on iPad. Seriously. If it looks broken in Split View or Slide Over, you’ve got layout assumptions baked in that will embarrass you when fold-state APIs drop. Check for hardcoded widths, fixed-size containers, and any place you’re using UIScreen.main.bounds (which, by the way, has been deprecated since iOS 16 — if you’re still using it, this is your sign).
Test with Dynamic Type cranked to the maximum. If your layout breaks at accessibility sizes, it’ll break at fold configurations too. Same muscles, same problems.
5. Fix your Swift concurrency warnings. Swift 7 is coming.
The Swift 6.2 migration made strict concurrency opt-in and more approachable. Swift 7, expected at WWDC, will likely tighten the defaults again — and the warnings you’ve been suppressing with @preconcurrency and nonisolated(unsafe) will graduate to errors.
This isn’t a guess. Apple’s trajectory is clear: every Swift version makes concurrency enforcement stricter while making the migration path gentler. Swift 7 is the “gentle but firm” release.
What to do this week: Open Build Settings, set SWIFT_STRICT_CONCURRENCY to complete. Count the warnings. If it’s under 50, you can knock them out in an afternoon. If it’s over 200, start with your data layer — models and services first, views last. The migration guide we wrote for Swift 6.2 still applies; the patterns are the same.
The one thing NOT to do
Don’t try to learn everything during WWDC week.
I’ve made this mistake every year since WWDC 2019. You stay up until 2 AM watching sessions. You start three sample projects. You rewrite half your app’s navigation using the hot new API from the Platform State of the Union. By Thursday, you’ve got five branches, no working code, and a caffeine headache.
Instead, use the next 26 days to get your house in order. Fix the warnings. Isolate the code that’s about to change. Read the reports. By the time Craig Federighi says “we think you’re going to love it,” you’ll actually be ready to love it — because you’ll be starting from clean code, not from technical debt.
Your one-page cheat sheet
| # | Item | Risk Level | Time to Fix | First Step |
|---|---|---|---|---|
| 1 | SiriKit → App Intents | High | 2-5 days | Search for import Intents |
| 2 | Core ML → Core AI prep | Medium | 1 day | Add a protocol wrapper |
| 3 | Liquid Glass audit | Medium | 1 day | Remove custom background hacks |
| 4 | Layout flexibility | Medium | 1-2 days | Test in iPad Split View |
| 5 | Swift concurrency | High | 1-5 days | Enable strict concurrency mode |
Keep learning
If you want to go deeper on any of these topics, our SwiftUI at Scale course covers App Intents, Live Activities, and modular architecture patterns that make exactly this kind of framework migration painless. The course was designed for moments like this — when Apple changes the rules and you need to move fast without breaking things.
And if you want to stay in the loop as we cover WWDC 2026 live, the blog is where it’ll happen. We’ve been tracking Siri’s transformation, Apple’s AI strategy, and the tools that are reshaping how we build iOS apps. The keynote is June 8. The prep starts now.
Mark Watney grew potatoes on Mars. You can run a grep for import Intents. Let’s go.
Share this post
Comments
Leave a comment
NativeFirst Team
EditorialThe NativeFirst team — engineers and designers building native Apple apps and writing the courses we wish we had when we started.