Siri Is Finally Getting a Real AI Upgrade: Everything Apple Is Expected to Show at WWDC on June 8

Apple has officially confirmed WWDC 2026 for June 8. The keynote begins at 10 am PT, streams on YouTube, Apple’s website, and the Apple Developer app. And for the first time in years, the whole tech world is watching Siri.

Here’s the thing: Apple promised a smarter Siri back at WWDC 2024. The ads ran. The iPhone 16 shipped. The features didn’t arrive. Apple pulled the ads, delayed the upgrade, and quietly settled a $250 million class-action lawsuit over it. That is the backdrop for everything happening on June 8.

This is not a routine developer conference. For Apple, it is closer to a public reckoning. Two years of broken promises, a CEO transition coming in September, and a global audience that genuinely doesn’t know if Apple can actually compete in the AI era. That tension is what makes this worth paying attention to even if you’re not an Apple user.

Here is everything expected to be announced, what it actually means, and what questions are still unanswered going in.

The Context Nobody Is Talking About Enough

WWDC 2026 is Tim Cook’s final keynote as CEO. Apple confirmed in April that John Ternus, head of Hardware Engineering and the architect behind Apple Silicon, takes over as CEO on September 1.

Ternus built the M-series chips that gave Apple the on-device processing headroom to even try building a serious AI assistant. Cook presents the software at WWDC. Ternus inherits it just in time to strap it to the iPhone 18 and whatever else ships in the fall. The software announced on June 8 is, in effect, the handover package.

That makes this keynote unusual. It is both a product launch and a statement of what Apple’s AI direction will look like under new leadership.

Siri 2.0: What Is Actually Changing

The rumor mill has been consistent enough across Bloomberg, 9to5Mac, MacRumors, and The Verge that the broad shape of the Siri overhaul is fairly clear. What Apple is building is not an incremental improvement to the assistant that has frustrated users for years. It’s a ground-up rebuild.

A Standalone Siri App

According to MacRumors, iOS 27 will debut a dedicated, standalone Siri app built around a chatbot-style layout. Think the conversational home screen you get with ChatGPT or Gemini, but native to your iPhone. It will show a history of your past conversations in a list or grid view, support both voice and text input, and include an attachment picker for sending photos and documents directly into a conversation.

This is the biggest structural change to Siri since it launched in 2011.

Dynamic Island Integration

Leaked renders from Bloomberg show Siri responses appearing as rich text cards that expand directly from the Dynamic Island. Quick queries get a compact answer right there at the top of the screen. For more complex conversations, you swipe down to open the full chatbot app. It’s a two-tier design: fast answers stay out of the way, deeper conversations get their own space.

A System-Wide “Search or Ask” Panel

Leaked details point to a new swipe-down panel accessible from anywhere in iOS, labeled something like “Search or Ask.” This puts Siri at the center of the operating system rather than the edge of it. The interaction would feel closer to how Spotlight works on Mac, but AI-powered and conversational from the first keystroke.

Chat History and Contextual Memory

Siri will reportedly remember past conversations. This sounds small until you realize that the current Siri forgets everything the moment you close it, which is one of the main reasons people stopped using it for anything meaningful. Persistent context is the feature that turns a voice assistant into something actually useful day to day.

The Google Gemini Partnership: What It Means and What Apple Is Being Careful About

This is the most consequential technical detail in the entire WWDC story.

Apple has reportedly used Google’s Gemini model to train a smaller, distilled version that runs locally on Apple hardware. The full Gemini model is the teacher. The compact on-device model is the student. Apple keeps the student, not the teacher, on your phone. All Gemini queries route through Apple’s Private Cloud Compute infrastructure, anonymized and not used to train Google’s models.

On Apple’s Q1 2026 earnings call, Tim Cook framed it carefully: “You should think of it as a collaboration. And we’ll obviously independently continue to do some of our own stuff.” He declined to share financial terms.

The strategic tension here is real. Apple cannot make Siri look dependent on Google without undermining its own identity. The framing Apple is likely to use is that Siri is still the personal, private assistant layer, and Gemini is one of the engines under the hood, not the product you interact with.

Whether users care about that distinction remains to be seen.

The Third-Party AI Model Switch: This Is Bigger Than People Realize

This is the detail that deserves more attention than it’s getting.

According to leaks ahead of WWDC, iOS 27 will let users choose their own AI backend. Gemini, ChatGPT, and Claude are all reportedly in the mix. This ends Siri’s exclusive OpenAI arrangement and opens Apple’s AI layer to genuine competition for the first time.

Apple already introduced ChatGPT as an optional extension in iOS 18, where Siri could pass certain requests to OpenAI with your permission. iOS 27 appears to expand this dramatically, potentially allowing any third-party AI agent downloaded from the App Store to plug into Siri via an Extensions framework.

iOS 27 test builds already show Apple’s own internal description of Extensions: “Extensions allow you to access generative AI capabilities from installed apps on demand, through Apple Intelligence features such as Siri, Writing Tools, Image Playground and more.”

That is Apple’s own language, not a rumor. The architecture exists. The question at WWDC is how they choose to show it to the world.

What this really means is that Apple is positioning itself not as the AI model, but as the AI operating layer. Your device, your privacy, your choice of model. That is a genuinely different strategy from anything Google, Microsoft, or Samsung is doing.

iOS 27: Everything Else Expected on June 8

Siri gets the keynote headline, but iOS 27 has more going on underneath it.

Photos App Gets Generative AI Editing

Apple is reportedly adding AI tools to extend, enhance, and reframe photos using generative on-device models. Think background extension, automatic perspective correction, and image cleanup, all without sending your photos to a cloud server. This takes aim directly at Google Photos’ Magic Eraser and Samsung’s Generative Edit.

Camera App and Visual Intelligence

Macworld reports a new Siri section inside the Camera app, making Visual Intelligence a front-and-center feature rather than a buried one. Visual Intelligence will also be able to read food nutrition labels and feed data directly into the Health app.

Safari Auto-Names Tab Groups

A small quality-of-life feature, but one that anyone who has thirty unnamed tabs open will appreciate immediately. Safari will automatically suggest names for tab groups based on what’s in them.

Smarter Keyboard Autocorrect

The keyboard is reportedly getting an autocorrect upgrade that suggests full sentence rewrites, not just single-word swaps. This moves closer to what Google’s Smart Compose has offered on Android for years.

Wallet App Adds Physical Pass Digitization

You will be able to scan a physical QR code or barcode from any ticket, membership card, or loyalty card and save it directly to the Wallet app. No dedicated app required for each one.

Accessibility Features Powered by Apple Intelligence

Apple previewed some of these ahead of Global Accessibility Awareness Day. VoiceOver gets more detailed descriptions of images systemwide, including receipts and photos. Magnifier gets an AI layer for describing content around the user. These are quiet but meaningful improvements for millions of people.

Performance and Stability Focus

Multiple reports describe iOS 27 as a “Snow Leopard” release, meaning Apple is cleaning up old code, removing outdated frameworks, and prioritizing underlying performance over flashy new features. If this is accurate, it could mean meaningful battery and speed improvements for existing devices.

macOS 27, iPadOS 27, and the Rest of the Lineup

Yanko Design’s detailed preview describes macOS 27 as a consolidation release, stable and foundational rather than headline-grabbing. The Liquid Glass design language introduced in macOS Tahoe (macOS 26) is staying, but readability improvements are coming to address the transparency and shadow quirks that frustrated users.

One detail worth flagging: macOS 27 is reportedly laying groundwork for touch-enabled Mac hardware. This is software preparation, not a product announcement, but it signals where the Mac is heading under Ternus’s tenure as CEO.

iPadOS 27 gets the same AI and Siri improvements as iOS 27, and watchOS 27 adds new watch faces including a variant of the Modular Ultra design. Nothing dramatic on the Watch side this year.

The Privacy Argument: Apple’s Real Differentiator

Every AI announcement Apple makes comes with a privacy layer that its competitors cannot easily replicate. Apple’s on-device AI approach means that for many queries, your data never leaves your phone. For queries that do require cloud processing, Apple’s Private Cloud Compute architecture runs on Apple Silicon servers, anonymizes requests, and does not use them for training.

The Gemini partnership created some tension here, since it involves routing some queries through Google’s infrastructure. Apple has maintained that even in this arrangement, queries are anonymized and not stored or used to train Google’s models. Independent verification of this claim will matter to privacy-focused users.

For the average user deciding between Android and iPhone, Apple’s privacy pitch on AI is still meaningfully stronger than what Google or Samsung offers by default. Whether that is enough to win back users who have drifted toward Gemini or ChatGPT for their daily AI tasks is the real question.

What Is Still Unknown Going Into June 8

The leaks have been specific, but a few things remain genuinely unclear.

Which iPhone models will support the new Siri. Apple Intelligence currently requires an iPhone 15 Pro or newer. If Siri 2.0 requires the same threshold, a large portion of iPhone users won’t see it immediately.
Whether Apple will demo the chatbot app live or only show renders. Apple’s track record with pre-announcement demos of Siri features has been rough. A live demo would carry far more weight than a polished video.
Exactly how the third-party model selection will work. Will it be a system-level setting? Per-task? App-by-app? The UX of model switching will determine whether most users ever change the default.
Whether new Mac hardware appears. There is speculation about M5-powered Macs, but no confirmation they will be announced at WWDC.
When iOS 27 actually ships. The expected window is September 2026 alongside new iPhone models, but Apple’s recent track record with Siri timelines means nobody is taking that for granted.

How to Watch and What to Do Right Now

The WWDC 2026 keynote starts at 10 am PT / 1 pm ET on Monday, June 8. You can watch it on:

Apple’s YouTube channel
Apple’s developer website
The Apple TV app on Apple TV devices
Apple.com during the keynote window

Developer betas of iOS 27 and macOS 27 will be available immediately after the keynote. Public betas typically follow in July. The full release to all users comes in September.

If you own an iPhone 15 Pro, 15 Pro Max, or any iPhone 16 model, you are already in the hardware window for Apple Intelligence. The Siri upgrades, when they ship, will come as a software update at no additional cost.

Can Apple Actually Catch Up?

Honestly, this is the only question that matters right now.

ChatGPT launched in November 2022. Google’s Gemini has been improving rapidly for two years. OpenAI’s voice mode is genuinely impressive. Siri, during this entire period, has been the assistant that people stopped trusting to set timers correctly.

What Apple has going for it is distribution and hardware. There are over two billion active Apple devices in the world. If Siri 2.0 is genuinely good, the rollout scale is unlike anything any AI lab can match. You don’t have to convince anyone to download an app. It is just there, on the phone they already carry.

The Extensions framework is interesting for a different reason. By opening the AI layer to third-party models, Apple is saying: we might not win the model race, but we can own the interface. If Siri becomes the layer through which you access Gemini, Claude, or ChatGPT, Apple stays relevant regardless of which model wins.

That is a defensible position. Whether the execution on June 8 matches the ambition is a completely different matter.

We’ll know in nine days.

If someone in your life keeps saying “Siri is useless,” send them this before June 8. It might be the last time that sentence is unambiguously true.

Sources and further reading

Apple, Tim Cook to become Apple Executive Chairman, John Ternus to become Apple CEO (Apple Newsroom, April 2026).

MacRumors, Apple to Pay $250 Million to Settle Class Action Over Delayed Siri Features (May 5, 2026).

MacRumors, Apple to Unveil iOS 27 and macOS 27 with These Rumored Features (May 7, 2026).

MacRumors, Apple Plans to Make On-Device AI a Key WWDC Focus (May 28, 2026).

MacRumors, WWDC 2026: Everything to Expect.

Memeburn, Apple Leaks New Siri: Looks Exactly Like ChatGPT (May 30, 2026).

Yanko Design, WWDC 2026: iOS 27, macOS 27 and the New Siri App (May 25, 2026).

FelloAI, WWDC 2026 Preview: Gemini Siri, iOS 27 and Extensions.

AppleInsider, Apple Doubling Down on On-Device AI at WWDC 2026 (May 28, 2026).

Macworld, WWDC 2026 Guide: Everything Apple Could Reveal.

TechRepublic, WWDC 2026 Preview: Apple Readies Siri Overhaul, AI Updates, and More.