WWDC 2026: Live updates from Apple Park on Siri, iOS 27, Apple Intelligence and more

If you have any questions about today's announcements, feel free to leave a comment below and I can go and ask Apple execs to their face. I'll be keeping an eye on the comments so keep it going! Thanks again for joining us today, you keep our journalism dreams alive!

Okay some of you have left some bangers that I have cackled at. I can't say which ones, but I'll just say I agree with you all. I'll shout out this one from jaswitzer97: "Pretty much everything expected from the bare minimum of what they promised YEARS ago, just repackaged; not really surprised by anything presented so far. We'll see if/when they deliver this time."

Right I've managed to find my way to the media area at Apple Park and am just starting to make sense of what's left to be done. We'll be closing this liveblog soon but I'll take a look at the comments to see if there's anything worth digging into! Thank you for joining us today, we appreciate you!

With that, we're done! Tim Cook and John Ternus were in the front row all the time and are currently being mobbed on their way out of this little theater.

That's the end of the questions and Federighi says he doesn't want to waste any more of our time. He cracks another joke about how many briefings we all have with Apple marketing folks after this and that's it.

Federighi adds that when things are sent to Private Cloud Compute (PCC), the data will be "vaporized" once the question is answered. "It's completely transient," Federighi said. None of the pieces of information that were sent off device will be stored to begin with. Both Rockwell and Federighi are adding that they can't look into PCC anyway, and they "can't even debug them," adding as an aside "hopefully that won't turn into a problem."

Rockwell brings up the drink-pairing demo from just now and says that was an example of what actually gets sent. Of the hundreds of thousands of messages that might be sitting on your device, only the two or so that are relevant were sent to World Knowledge, for instance to get the best wine-pairings. None of the information about who sent the message or when it was sent went to the cloud, either.

Next: "You keep mentioning that only information needed is sent to the models. What actually is "information needed"?" 

Anyone of us can answer that one, Federighi said.

Federighi uses his words and talks about Xcode serving as a possible surface where agentic AI appears. "It's all about finding the right user experience, one that the user can understand and control" and that's safe, Federighi says.

When people hear "agents" they might think about things like OpenClaw, Federighi says. He throws to Rockwell to talk about Siri's "agentic foundation" and Rockwell is explaining what an agent is. "Ours is primarily request-based today," he says. "Our ability to extend to the future, is.." he trails off but nods in a vaguely positive vibe.

The next question: "We didn't hear much about AGENTS!" Federighi remarks about the use of all caps and jokes "you guys must have been at [Google] I/O."

A second question is about the balance between personalization and privacy when it comes to the power of Siri AI. Sebastien is answering it by saying that being able to get personalized answers is very powerful. Rockwell jumped in to say that personal data is stored on device, and when you send a request, only the pieces that are useful for that moment will be sent to the cloud. 

Federighi answers it more by saying it's being conscious and "you're not going to see an ad for us" about Thai food if it thinks you like Thai food. If you see a recommendation for restaurants, it won't be because it's serving ads but because it might actually be relevant to your preferences.

The first question is about what caused the approach to Siri to shift and become something that seems more chatbot-like. I have to admit, the question was very long and worded meanderingly so I might have gotten it a bit wrong.

It will take "programming on our platforms to the next level," Federighi said. He's moving on to questions submitted after the keynote from this specific audience in person here.

Federighi is talking about how Swift got a huge update this year, which is something we already learned about at the keynote.

Federighi is now talking about how the Spotlight semantic index is open to third-party developers, which can then enable users to find content within apps.

"We don't want people second-guessing, when they're invoking Siri, whether their data is going to remain private or not." This is particularly important if Siri can see everything on the user's screen... Which is reminding me of Microsoft's very humiliating Recall debacle.

Sebastien says if Rockwell were to sent a message inviting him to a concert, Sebastien can just ask Siri to tell him if he's free that day.

Siri gives a long answer that mentions Mount Tamalpais, and Sebastien asks it to create a plan for a trip to "Mount Tam." I guess that's what the cool Bay area people call it?

Sebastien is going to "follow Mike and try a live demo" to give us a little context. He picked up an iPhone, which has a photo of a grassy landscape on it. He triggers Siri and asks "why are the clouds like this and where can I see this in the Bay."

"This model is really exciting" and is really beginning and allows Apple to build on the strengths of the architecture, Federighi takes over to say. And brings us back to "what's really unique with the new Siri." One of the elements, he says, is onscreen awareness, and he throws it to Sebastien (whose last name is really difficult for me to type accurately and quickly.)

The models have changed how Apple was able to train these new voices, Subramanya concluded. Rockwell jumps in to say these models allowed Apple to use 20-billion parameter model and depending on the request it's using anything between 1 and 4 billion parameters. The number of times "billion parameter model" is being said here is really too high for me right now.

He's describing a situation where systems have to juggle requests for tokens and how that is more deftly handled on servers than smaller devices. This is why Apple built a sparse model for on-device use. I don't think you need to know about the need to reload parameters per token, do you.

Federighi is back at the helm and referencing the voice we heard in the demo by Rockwell. He's asking Subramanya to talk about how the voices for Siri are handled by AFM.

After the user asks the question, the system orchestrator will decide what needs to happen. In the instance of "what drinks would pair well with those two dishes" the system orchestrator has historical context of what "those two dishes" were but can determine it needs to tap world knowledge for the answer.

Then Rockwell asked Siri to suggest good drink pairings with the food items people are bringing. Rockwell has stood up to run us through the process that happened during that demo, with the chart we saw earlier.

Rockwell asked Siri on the phone in his hand to tell him what everyone was bringing to the potluck party he was planning and after a little chime, Siri displayed the answer in an expanded Dynamic Island window. A new Siri voice also read out the response, and you can see in the interface where the assistant was pulling information from. In this demo, it was from two separate chats in Messages.

"All good for me to talk about that stuff, but maybe I thought I would show you the new Siri in action," Rockwell said. Live demo time!

One of the more powerful aspects of Siri is the ability to ask about your personal context. Again, we've heard this all before.

The result is a "profoundly more capable Siri," Rockwell said. It's got its own app, it's natively multimodal and has privacy built in from the ground up. It works in all Apple products like mobile, CarPlay and more.

The vision Rockwell and his team had was to make Siri the most powerful and capable assistant possible. There was a working version last year, Rockwell said, but "we didn't feel it was delivering" on the vision and experience it was supposed to do. So the team went back and rebuilt it "from the ground up" on top of the models in AFM we just learned about.

Another theme that "ran through many of the questions," Federighi said, was about Siri. Are we getting it finally? What can it do? The answer to that first one is a resounding yes! If you install the beta. Mike Rockwell is answering the question by first talking about how exciting of a moment it is for him and his team.

Apple is in charge of what gets deployed to these nodes, Sebastien said. Apple devices themselves are only able to talk to software that has been signed by Apple, so even if they're talking to third-party cloud processors, unsigned parties can't access or communicate with those devices.

Sebastien reiterates that privacy is at the core of Apple Intelligence, and that Apple has set out to extend Private Cloud Compute to work with third parties. It's not very clear to me at the moment what has really been done other than "collaboration" and I'm waiting for the answer to become clearer.

The next question has to do with whether the partnership with Google means user data is also shared with the Search giant. Sebastien is answering this. 

For tasks like complex reasoning, there's AFM Cloud Pro that Subramanya said is able to handle more-demanding image generation tools. 

One of them is the AFM Cloud, which is the "server workhorse model," and is optimized for latency. There's ADM and Cloud Image, which is the next generation of the company's image editing and generation tools, including things like Spatial Reframing. There were also AFM Core and Core Advanced, and they were all built, together with Cloud, for Apple Silicon devices. They were refined with Gemini models.

The third person I hadn't yet named, by the way, is Amar Subramanya, VP of AI at Apple, who is currently naming the three models in the AFM family. 

They're about to answer questions that were submitted by the audience and are starting by addressing the one about the partnership between Google and Apple on the Foundation Models, which is in its third generation (and referred to as AFM).

Federighi brings out members of Apple's various teams like Sebastien Marineau-Mes, Mike Rockwell and one other person who all joke they all report to Federighi and that's why they all "wear the blue shirt."

Finally when you ask about real world events that require world knowledge, they go to Apple's World Knowledge databsse.

"In that deployment model, we have a family of third-generation Apple Foundation Models" he added. These models are the product of the collaboration between Apple and Google, and they're architected to run on Apple's own architecture. They were designed specifically for Apple's ecosystem.

When more power is needed, the system will tap Private Cloud Compute. Federighi reiterates all of the requests are private and not accessible by Apple or third parties. 

Next is the "system orchestrator," which is key to the whole process, Federighi said. This orchestrator directs processes through the App toolbox, Spotlight index and on-screen content. This is all built on top of a set of "powerful on-device models."

As he describes these steps the graphic is adding more information and I'll be sure to share that once it's more complete.

"Plugged into that is the Siri App," he said. 

"Let's talk about what we do use or how our system is built." He said it starts with the "system experience" meaning how the Siri experience is build into iOS, iPadOS etc.

He's acknowledged that Apple uses "none of those things" — things like client codes or models that Google deploys to its customers. None of the same infrastructure or knowledge base that Google uses either. "So, uh, hope that's clear," he said.

He's showing a graphic that demonstrates the "traditional chatbot architecture" and how a chatbot app might interact between a device and cloud-based models.

"One of your big questions that we've seen so far have been about Siri and its architecture," he said, especially about its partnership with Google's Gemini.

He's talking about the Reframe feature, but mostly recapping what it does. He's also brought up Visual Intelligence and writing assistance with Siri. Nothing new just yet.

Federighi is retreading some of the topics covered earlier, recapping the announcements. He also said earlier that there will be answers to some questions that were submitted by the audience live through a (private) online portal.

I guess Cook's attendance was the cue. Federighi has come out on stage and acknowledged that there was a lot of news announced just an hour ago, and that there have already been a lot of questions asked by people on social media.

No one has taken the stage yet but Tim Cook has entered the auditorium.

A disembodied voice has come on overhead to ask us to note the nearest emergency exits, so I think things are kicking off soon!

I wonder if I should just get Visual Intelligence and Siri to do it for me.

The music has just died down here but there's still about 5 minutes to go. It's really dark here so I'm not sure I can tell if he will be in a perfect shade of light blue but I'll do my best! My eyes are old, y'know.

Looking forward to that, also please update us on if he's still wearing a blue shirt.

In about ten minutes this secondary event with Craig Federighi is going to begin, so stay tuned. I'll be telling you what it's all about as it happens!

Meanwhile, I have relocated back to Apple's Visitor Center where there is, thankfully, AC and much better Wi-Fi. One thing I'm still thinking about is that spatial reframing demo towards the end of the keynote. I know photography purists will have a lot to say about this feature. But it is an interesting example of how Apple has used learnings VisionOS in other products. In this case, it enables Apple to understand where subjects of photos are in relation to their surroundings so the image can be completely recomposed. I'll be very curious to see just how smooth it is in practice, though.

Here we are at undisclosed secondary location for a technical deep dive with Craig Federighi!

I'm currently walking with a group of media to a secondary location for something titled Tech Talk with Craig Federighi and I will likely liveblog it!

The keynote was a lot shorter than I expected but we have plenty more to say... But I'm heading off to my next location now while Karissa heads to the bathroom for a well-deserved toilet break. We're still here though, and I'm about to go look at the comments you all left!

That's 12PM PT or 3PM ET, and given how surface-level quite a lot of today's announcements were, I'm sure I'll have more actual details to share later today.

We'll keep things running here for a bit, but I have a mysterious 12PM placeholder where I might learn more about today's announcements, so stay tuned. 

And that's a wrap on WWDC 2026. It was... all about Siri and Apple Intelligence. There was no dedicated moment for watchOS for example, or something about iPadOS. Felt quite different than usual!

And we're out, I wonder if we will see John Ternus under that same rainbow at next year's WWDC.

"I truly believe the best is still ahead," Cook says in this video. He talks about how Apple's north star has always been creating products that serve people's needs. It's been an honor, he says, before thanking everyone for joining the company today.

Over the years, you have helped people connect, create, learn, Cook says to the developers here. 

The new OS releases will be available as developer releases today, and will be available to public users this fall. On a personal note, Cook notes, some of the highlights has been events like this.

Cook thanks the presenters and the developers gathered today. "The next generation of Apple Intelligence introduces new capabilities including Siri AI," he says in the pre-recorded video we're all watching.

Tim Cook is back to close out his last WWDC as Apple CEO. 

Developers can easily resize and interact with app previews, and they'll learn more in upcoming sessions and state of the union presentations.

Craig says Xcode is the "best place" for agentic coding and, um, I bet a lot of developers in this room might have something to say about that.

"We're reimagining how you can test apps with the all-new device hubs."

"Our app ecosystem has never been more vibrant." Federighi says agentic coding is changing how apps get built. Xcode has a coding assistant that can simulate entire apps. 

Federighi is talking about how AI uses its... fashion expertise to find the right outfits for his multiple keynote presentations while staying within his "narrow fashion comfort zone" of blue shirts and blue shirts alone.

Developers can use images as input in the foundation models framework now. Custom skills and server models are available to developers.

Federighi is describing how some third party apps like Line can use App Intents to allow users to ask Siri to do things on their behalf.

The most powerful on-device models are coming to most recent iPhones and iPads, and the new Siri is ready for testing for developers later today. It won't be available in the EU immediately, though. In China, Apple is working through regulatory requirements so it also won't be immediately available there.

A big boo in the crowd when Craig says that Siri AI won't be available in EU. 

Federighi says this is "a big step forward on our journey to integrate powerful AI" into all the company's platforms. More to come, too, which I'm sure I'll learn about in detail across different briefings later today.

Some features require paid access (which are included with some iCloud+ tiers).

Federighi is back and recapping the new Apple Intelligence features, and they'll be coming for free in all the languages supported currently by Apple Intelligence.

The new features will work on older photos and those taken with other cameras, too.

This is all done on Apple's Private Cloud Compute so while it does get processed in the cloud, your data shouldn't be compromised. The Reframed image takes a few seconds to show up, in this video, but it does look solidly... reframed.

In the photo editing toolkit, selecting Reframe allows you to interact with the scene, and there is a blur around the edges of the original image which "will be filled in later" with Apple's generational AI.

Deshpande is showing an example of a photo of his children in what appears to be a front yard.

The last one builds on Cleanup and Extend. If you've ever wanted to reframe a picture after taking it, this new tool called Spatial Reframing can help.

You can use the new Extend tool to add more background space or adjust the aspect ratio of a picture. 

Apple Intelligence will help edit photos in three key ways. Cleanup is getting a "big upgrade" to remove distractions with more realistic infill "even when the scene is complex." 

The goal is to help photographers in ways that "respect the original moment." 

AI photo editing.

Alok Deshpande is up to talk about Apple Intelligence in the Photos app. 

You can start an invite from the Photos app by selecting a person and then describing your event. You can then use Apple's foundational models to generate more types of images and use them as things like Message backgrounds or lock screen wallpapers. Developers can make use of the new Image Playground API.

The new Apple generative model now allows for photorealistic styles in Image Playground, which used to be limited to the ChatGPT-powered model. You can also describe what you want more naturally and Apple Intelligence will better understand you.

Next up is Leslie Ikemoto, who is talking about how Apple Intelligence is integrating into communication tools and expression features like Image Playground. 

Shortcuts is one of those features that can be super powerful, but has always been more of an "advanced" level thing. Glad to see Apple making this a bit easier to use.

If you want to let your partner know what time you'll be home every time you leave home, just say exactly that to the Shortcuts app and Apple Intelligence will do the math in the background based on your ETA and location, and then send the right time to your designated recipient.

The Home app can do Amazon-like analysis of your camera feeds.

Shortcuts is also getting an Apple Intelligence update. It will be more approachable thanks to Apple Intelligence, which can help turn your natural language descriptions of a shortcut into multi-step functions.

You can then put sequences of events together thanks to Apple Intelligence. It will also surface clips you should review at the top of your app. If you're looking specifically for an incident, like a package delivery, Apple Intelligence can surface it for you across various cameras.

Apple Intelligence can group notifications into updating alerts for single activities. The Home app can also use footage from compatible cameras to generate descriptions for what happened in your surveilled areas.

Over to Cecilia Dantas, who is talking about how Apple Intelligence can "streamline the things you do everything." We're starting with Home app notifications.

Apple Intelligence also powers Call Context in the Phone app, so if you call an airline to change your flight, for example, the Phone app can find your confirmation code in Mail. This is nearly identical to Magic Cue in the Pixel 10 phones by Google.

Calendar also allows you to add events to your schedule via a natural language description, instead of having to tap copious fields and such.

Messages will use Apple Intelligence to understand the content of your messages and suggest actions. Like creating reminders or finding relevant photos when your conversation suggests you should. This is all very similar to Google's Magic Cue.

Next up, Passwords. Most of us have had passwords compromised by now, and Passwords can not only tell you when a password has been exposed, but also automatically change and update them for you.

Next: personalization. If you could add a button to a toolbar for easy access to a specific part of a specific webpage, wouldn't it be great? Describe an extension is a new Safari tool that can do that.

Safari doesn't share your browsing data with Apple or anyone, to keep your data private.

Sometimes we leave tabs open to wait for something to be ready. Safari can automatically monitor your tab on your behalf with Notify Me. You can tell Safari what you're looking for and close the tab. When the change has happened, Safari can send you an alert.

Safari can add new related tabs to a topic "even as you browse." You can close them after you're done browsing or save them as a Tab Group.

Tab Groups can be helpful but it might not be enough. Now, Safari can "use Apple Intelligence to tame your tabs by organizing them into topics."

Apps like Passwords, Safari, Messages and more will be getting Apple Intelligence updates. Safari, for instance, is getting simpler ways to manage tabs.

From browsing to editing photos to composing messages, Apple Intelligence can do a lot more. Federighi hands off to Beth Dakin to talk more about the details.

Siri AI will be available in English to start, with support for more languages coming "quickly." Now, updates on Apple Intelligence.

Back to Federighi, who is back recapping what we've learned about Siri AI. Fun fact, this is probably the most that Apple has said "AI" in a WWDC or keynote ever. 

Siri can give tips and suggestions on how to improve your writing after you highlight it. Apple Intelligence will automatically proofread for you, and it's available systemwide, including in third party apps. I'm not sure I like that.

Highlight the text you have in a document, for instance, and you can get help from Siri when you draft an email with it.

Wow, Siri getting some multimodal capabilities. I wonder what else you can do besides the "does this fit in my suitcase" example. 

Visual Intelligence in Siri is also coming to visionOS, he says. Siri can recognize what you're looking at and tap realworld information for answers. You can also ask Siri for details on real objects around you. In addition to Visual Intelligence, writing tools are also getting an update.

This feels a bit like what Google has done with Circle to Search on Chrome – being able to get context and answer based on what you select on your screen.

On macOS you can select areas with the screenshot tool to get more answers and details on. If you're looking at a schedule, you can bring up Visual Intelligence and quickly add all the performances in a festival to your calendar, for example.

Sebastien Marineau-Mes is talking about Siri mode across other apps like Camera. It has image understanding that's powered by Apple Foundational models. You can point your iPhone at a plate of food "to get nutritional insights" or take a photo of a bill and use the AI to decide how to split it.

New and improved Siri on visionOS. 

"We're also tailoring Siri AI for watchOS" and there is a new app grid on Apple Watch. Siri AI is also coming to visionOS, where there is a new 3D visualization of the AI. You can just look at the orb to interact with it, rather than having to say "Hey Siri."

Across Apple products, the new Siri app "makes it easy to revisit an existing conversation or open a new one." You can see conversation histories there, which are "synced privately with iCloud."

He's asking the assistant about a problem his child had with one of the setups and asks Siri what he can do about it. The assistant found the message from his child, found a possible solution to the problem and, at Justin's request, composed and sent an email to the right contractor.

Siri returns a text-heavy response, followed by a three-column table with details on how to adjust his shed. Justin has different quotes in his hard drive for different sheds. He selected a few of them and brings up the context menu where there is a new Siri field. He asks it to "compare these" and the assistant returns a table.

The new Siri experience in Spotlight macOS.

There are system-wide context menus, too, and Justin is going to give us "a closer look." He opens a picture from a folder on his Mac, for example, and asked Siri via Spotlight about how to build a "maker space in the shed."

Justin says he can refine the menu further, but for now, he wants to share it with his friends. He asked Siri to share it with his friends and "include the menu" and the system drafted a message and shows him buttons to send or edit the message.

Justin asks Siri about a dessert from an event he was at recently, and the assistant pulls the information from his messages. He tells the assistant to "bring this together" into a menu for his watch party.

The new conversational interface for Siri, it can also pull in info from your phone like Messages.

Justin is asking about the World Cup, and then getting information about dishes from specific countries participating in the sport tournament (I know what the World Cup is).

You can ask about an in-depth plan or get feedback on documents. These will be available systemwide and also get answers from anywhere in the system. You can drag down from the dynamic island or long press the Action Button.

Voice dictation, built right into the keyboard, is also getting updated with improvements to capitalization and more. These will be coming to things like CarPlay and elsewhere. Up now is Justin to talk more about Siri AI.

You can make micro-adjustments to Siri's voice.

We're watching how a custom voice can be created in the Siri setup. 

"These are just the ways that Siri AI are much more capable" Rockwell said and he is introducing a new voice experience. "Joz said these apricots are peak, why has no one ever told me?" and "Clear the fridge. No clear all the fridges." And the AI voice certainly expressed it all with seemingly appropriate intonations.

"Add just the one with Bryce, Madison and Quinn to our shared family album," Rockwell said, and of course the assistant did that dutifully.

Rockwell showed how Siri can understand what's on the screen, answering questions about landmarks in pictures on the display, and delivering navigation guidance to specific locations. Rockwell asks it to show photos from a recent family trip to a ranch.

A little Siri demo that shows the new UI. Fun fact: I'm actually in the lottery for this show later this summer. 

Now we're watching Siri pull up music by Waterhouse. Next, Rockwell is showing us how Siri can understand what's on your screen. I have to wonder how much of this pre-recorded livestream is.. real. 

It also works with Visual Intelligence and you can ask it to write for you. There is a new design, too. We're going in for a closer look now. Rockwell is showing off how Siri can draw on current world information to get your answers. A demo plays of him asking about an upcoming Suki Waterhouse concert tickets. After Siri informed him that he has to enter a lottery for them, he asked it to "remind me to sign up when the lottery happens." It set a reminder.

Siri is now more capable, more conversational and with the dedicated Siri app you can refer to previous conversations.

It's Siri AI.

"We've rebuilt Siri with powerful AI at the core and today we're introducing an entirely new version of Siri," Rockwell said. "We call it Siri AI."

Apple isn't like that, Federighi says. Your data is only used to execute your requests, and Apple has allowed third-parties to validate that it's doing so. Next up, the all-new Siri. Federighi has handed things over to Mike Rockwell, the VP of Siri engineering.

Privacy is another area where Apple clearly thinks it has a big advantage over competitors because it doesn't access users' personal data the way a lot of other companies do.

Because they're systemwide they're more useful for the things you do every day, Federighi said. Today, many AI systems hold on to your information and rely on you to defend your privacy by using temporary chats or deleting your information manually.

There's also onscreen awareness to tailor its assistance in the moment based on what you're doing. This is all sounding familiar to me because I feel like I heard the exact same thing two years ago?

Behind the scenes, Apple Intelligence uses Spotlight and its "powerful semantic index" to understand what information is available to it. It can also go to the web and use cloud compute to generate an answer when needed. App Actions also allow Apple Intelligence to use tools from apps to complete your requests.

A second, more powerful version of the on-device model of Apple's ML has been designed for devices with Apple Silicon that can make voice recognition much better and enrich typing experiences. Apple Intelligence has a new system orchestrator across the company's products, too.

First up, the architecture. It's centered, again, around you and the things you use every day. "We embarked on a deep collaboration with Google" this year, Federighi said. Together with Gemini models, Apple's own models have been developed to work on device and on servers. They unlock huge upgrades and multimodal capabilities, he added.

We'll learn about its new architecture, new Siri and how it makes apps smarter and more useful.

Did you notice the subtle dig about other companies pursuing AI "just for the sake" of it? Amusing coming from Apple, which has been fairly late to the game here.

Apple believes that truly helpful AI must be centered around "you and your needs," he said. This means AI integrated deep into the products you use every day, designing it with privacy at every step, he said. Apple Intelligence is up next.

Federighi says AI is incredibly powerful technology that, with proper care, can "unlock meaningful benefits for people everywhere." But some appear to be racing forward to pursue AI for the sake of AI without care or concern for the people it's meant to serve, Federighi said.

Declared age range has been a thing for awhile, but I expect it will become even more important as we see more and more age verification laws come online.

Developers could use a child's age range to tailor their experiences accordingly, for instance. Back to Federighi now, who is summing up the teen and kid feature updates.

What about experiences kids have within apps? Apple is aware parents want to make sure their kids are ready for things like AI or in-app chatting. Thai says it's developers that play and important role here on shielding kids from content they shouldn't see. Apple believes developers should shoulder that responsibility and says it's setting up guidelines to help developers to create those frameworks.

I know parents have been begging Apple for years for better and more nuanced screentime and parental control features. My big question is whether these will be harder for kids to circumvent, I often hear that the existing controls are incredibly easy to get around.

Weekends and week days are different, Bose jumps in to say. He introduces Screen Time Schedule to allow greater flexibility and awareness of days of the week. The schedule can work "hand in hand" with time allowances so kids can request more movie or game time on a Saturday, for example.

These suggestions are based on a child's age and will get updated as science evolves and parents share feedback. When it comes to social media, experts recommend children under 13 don't use social media and that parents carefully consider when their kids are ready, Thai said.

Screen Time is up next. Parents want to better manage how much time their kids spend on specific apps. Time Allowance will put entertainment, games and social media apps front and center with daily recommended limits for all three. 

As for who kids can talk to, parents can set up a list of people like grandparents or relatives to allow access. Kids can also ask for permission to add new people to the circle. There's also been features like Communication Safety which intervenes by blurring out potentially sensitive content. This year, this will expand to include gory content.

There are age-appropriate labels in the store, and kids can ask for permission from parents before installing them. This year, Apple is adding "Ask to browse" as a way to let kids to seek permission for browsing content on the web via Safari, and it's on by default for kids under 13.

Bose just handed off to Ann Thai to talk about managing which apps kids can download from the App store.

Starting with the content kids can see, Bose is talking about why it's important for children to have their own devices. But it's important to prevent opening them up to too much too quickly. Starting this year, parents can "start focused" and "add more over time." It "begins with an easy-to-use setup assistant" that walks parents through the process.

Bose is recapping features that child accounts offer, like parental controls and limits set on App store. This year, Apple is focusing on the content kids can see, who they're allowed to talk to, when they can have access and how to guide them on their "digital journeys."

Apple has developed a guide and continues to collaborate with researchers to advancing the science in this area. Dr. Desai hands of to Raja Bose to talk about how parents can use Apple products to set up products for children.

Dr Desai talks about how setting up boundaries for technology use is critical for kids developing heathily. Research shows that kids under 18 using personal devices benefit from age-based protections guided by adult supervision.

These features are especially important when it comes to kids and teens, he says. As parents themselves, people at Apple are committed to expanding child safety features "with powerful and intuitive tools," he said, and hands off to Dr Sumbul Ahmad Desai.

He recaps features like end to end encryption in messages, Safari privacy as well as "life-saving features like crash detection."

Ford hands off to Federighi who is recapping (in a strangely booming voice) to talk about trust and safety.

Spatial Scenes on visionOS are getting an update to be more immersive as well, and Maps is getting a boost in Flyover to improve the detail and quality of the images and footage.

The Health app will now offer support for tracking perimenopause and menopause. As someone who has been getting a lot of hot flashes lately this is music to my ears. Speaking of, AirPods are getting custom EQ capability as well.

Big applause just now for being able to share full-res photos in shared albums.

Those are just "some highlights of the work we've done this year," Ford said and alludes to some "great features" across other products. For example, iCloud shared albums in Photos will now support Android and Windows users. It'll also work with full-resolution photos.

The index has been rearchitected to be more stable, efficient and comprehensive of old and new content. After you update, the new infrastructure will get started reindexing. New content will be indexed "almost immediately," so your stuff will be searchable sooner.

Search is getting a significant update, it seems. On iOS, iPadOS and macOS, the foundation of Search that powers Photos, Spotlight and Mail has been rebuilt.

It's also working on clearing speed bumps. Network transitions, for example, will get more seamless. If you're switching between cellular and Wi-Fi, it should happen autonomously more quickly. Or if you're going off a plane and still stuck on the onboard Wi-Fi, it should not stay stuck as long on the new iOS. The Message app will also show a new progress bar to indicate the progress of your message relay.

iOS 27 is supported on iPhone 11 and all the same iPhone models as iOS 26, Ford said. This means iOS 27 is available to more users than any iOS release ever.

Older iPhones are getting updates to the CPU scheduler that will make them feel faster. On newer phones and versions of iOS, the scheduler has been optimized as well.

As someone currently struggling to get AirDrop to work, I welcome any and all improvements.

A lot more things "are faster" and Ford is naming just "a few:" iOS apps will launch 30 percent faster, for instance. Content across the system will also load faster. After you take a picture, new photos will appear in your library up to 70 percent faster than before. Sending them through AirDrop will also be speedier.

Back to Ford, who says these "design improvements are just the beginning." Fundamentals like CPU usage and memory management were also looked at. "We optimized the parts of the system that make a big difference in the responsiveness of our products." System animations, for example, have been updated to be faster.

Every window on macOS will now have the same tighter corner radius, and app icons will be getting additional layers of Liquid Glass directly. These should help make the graphics look sharper and more defined.

In addition to improving Liquid Glass, some of the "cornerstones of macOS design" will also be getting updates. A new more uniform toolbar is coming to all apps, and it should offer better structure. Sidebars now expand to the edges of the screen, too, to reduce distractions at the corners.

The team appreciates feedback on new design languages like Liquid Glass, Kedia said. It's updating how Liquid Glass looks, starting with improving readability. But in addition to tweaking the default look, Apple is adding a new slider in settings to adjust Liquid Glass. It can go from ultra clear to fully tinted.

Ford is starting with Design as one of the key themes of the effort across Apple's paltforms and hands things off to Shubham Kedia. Kedia recaps last year's Liquid Glass rollout.

Federighi hands off to Stacy Ford, VP of OS program management.

It was some sort of weird poem-haiku-hybrid. And a Volkswagen van pulls up with a bobblehead figurine of Federighi on the dashboard and some kind of smoke/incense behind it. The name of the next macOS? Golden Gate.

macOS appears to be first up. Federighi is talking about how macOS' naming and marketing teams might have gone rogue instead of sharing the actual name with Federighi. He is now reading a note he claims was passed on to him by the macOS team.

All platforms, like iOS, iPadOS, visionOS, tvOS and macOS, will be covered today.

Cook hands the "stage" (in the video) off to Federighi to get us started. "This year we put our focus into three areas," he said. They're improvements to platforms, updates in trust and safety and finally "a big leap forward for Apple Intelligence." It has "an innovative architecture that unlocks a new Siri across platforms."

Today, announcements will be across Apple Intelligence and Siri, among other areas.

Cook is now going over the ways developers can build experiences across Apple's entire ecosystem of products.

"WWDC also gives us the chance to celebrate our incredible global community of developers," Cook says. There are well over 1,000 submissions to the App store every hour.

Tim Cook appears... in the video. And is welcoming everyone to Apple Park and WWDC. 

"We've got a huge day ahead, so let's get on with the show!" And we're off with a video that no doubt many of you will see at home too.

He's thanking the audience for being with him on his long career at Apple. Developers are a core part of the reason Apple's products are as powerful as they are. Cook also thanks members of the Apple team for making "today's announcements possible."

"Thank you, thank you, thank you," Cook has been saying for the past minute. "I've never seen so many iPhones before."

People are all stood up, but not necessarily giving an ovation. At least half of them have their phones up.

Federighi is talking about what Cook's tenure has seen. "Tim loves developers. He's never happier than when he's sitting down at Apple Developer academy." And.. he's talking about how when he first met Cook he was "just a kid off the mean streets of Cupertino." And Tim Cook is walking out on stage now to lots of cheering.

He's talking about how this is the last WWDC where Tim Cook will be presenting it as Apple's CEO. It'll be Cook's 15th WWDC.

Craig Federighi has just stepped out onstage to say hi to the audience here in person.

Right we're just about 7 minutes out from the keynote starting and I'm pleased to announce that the sun (or really, the earth) has moved enough that we're in the shade and my laptop will not melt. Bathroom break, check. We're ready to go. 

Traicovn I see you in the comments section with this biting take: "Perhaps nothing has ever been more fitting as an analogy for Siri than a 'breakfast hot pocket'. Ooooh! This looks nice, seems hot. Oh wait... no... this isn't fully cooked, it's still frozen in the middle."

I can't wait to see what gets announced today and whether the analogy still holds!

We just got the 10 minute warning for the keynote start. Time to refill your coffee, grab some snacks, go to the bathroom and whatever else you have to do. (I'm reapplying my sunscreen.)

I've been taking pictures and posting from an iPhone 17 Pro Max, by the way, if you were feeling like critiquing the quality of my photos.

I'm desperately waiting for the sun to move enough so that we're no longer being slowly roasted to a crisp. If I'm not directly in sunlight, it's actually a beautiful day here in Cupertino!

We've just gotten to the seating area and I think we've nabbed a good spot!

I have to say that one of my favorite trends of the last few years is more companies moving their keynotes outdoors. It can be rough if you have a seat directly in the sun, but it's so much nicer to be in the fresh air instead of over-airconditioned convention centers where you go all day without seeing the sun.

The seating area in front of the stage is filling up!

An Apple TV refresh would be very welcome in my household ngl. I feel like every WWDC I can remember has "better Siri" on the agenda and we've mostly seen incremental improvements. If they can't get it together this year, then I'm not sure there's much hope. That said, I think after last year's debacle we have good reason to be (very cautiously) optimistic.

RECL left a hilarious (and true) comment: "with all the spoilers online this events turns into a rumor checklist"

I mean I always want to check what percentage of rumors actually turn out to be accurate after every major event, and I haven't had a chance to do it!

Troy said in the comments: "I want the Apple TV also and am sad that it is waiting on the new Siri. But I am happy for a new Siri!"

Looks like we have a lot of Apple TV users in our reader audience! It's going to be nice to be able to ask Siri to play something for me and jump to specific scenes. At least, that's what I'm imagining a new feature would be.

Okay I'm going to take a moment to answer these comments that have accumulated!

Huss asked: "Is today only going to be about software or will we be getting hardware announcements too (my TV is begging for the updated Apple TV 🤞)?"

I am praying there is no hardware because it will be less work for me but we're certainly preparing ourselves to have someone write about tvOS today. Does that help? There's usually not much hardware news at WWDC!

At the risk of pivoting too hard and posting too much about the in-person vibes instead of real news... I will tell you all what's on the breakfast menu today:

• Breakfast hot pockets (with vegetarian version available)
• Greek style frittatas
• Appe pie and pecan overnight oats
• Banana smoothie with nilla crumble
• Plum and amarena cherry fruit salad
• Mini croissant
• Mini creme feuillete

They're all perfectly sized portions, in my opinion!

That's a pretty serious assertion, and I wonder whether Apple will get around to addressing it. Not today, for sure, though.

Good morning! I've made it into Apple Park. A bit more of a spectacle than usual this morning as there were protesters from Ultraviolet and Heat Initiative outside. I was escorted past them fairly quickly, but they were talking about child safety issues. (I believe they've mounted similar protests at other Apple events.)

We're hearing that some sort of protest is happening here today. Karissa has a picture, and my friend Julian Chokkattu from Wired says he heard someone (not an Apple staff member) at registration area claim that their daughter was exposed on the Nudify app...

Karissa is here! The parking was far away!

I have just arrived at the media reception area and will be prioritizing coffee for a bit.

We have to walk a fairly long way to get to the stage and seating area but at least it's pretty and nice out! 

Received my badge!

Video is up and I'm next in line to check in!

I have arrived and am in line for media check in! Am about to post a video of the long lines of developers to our Instagram Stories so check it out there!

I hear Karissa is in a long line of cars heading into a parking garage and she's being ushered by people wearing high visibility vests. Karissa if the parking garage looks like the basement of a Severance office then you're in the right spot. And you'll never get out...

I'm about to make my way to Apple Park to pick up my badge and check in (we will only be allowed in at 11AM ET) — wish us the best traffic luck! Meanwhile, if you haven't already, I encourage you go head on over to the Engadget social media channels (particularly Instagram, if you have it) for additional behind-the-scenes vibes!

One thing to note — you might have seen online many people posted pictures with Apple's incoming CEO John Ternus last night. We were at a welcome event for media, and Ternus was basically mobbed the entire time with people clamoring for selfies and photo opps. So while I can confirm that those pictures are not AI, I can't post my own selfie (I have video, though!)

In case you weren't aware, thanks to the folks at Bloomberg (and elsewhere), we already have a pretty decent idea of what to expect later today. The publication mocked up some graphics of what it expects the iOS 27 interface for Siri might look like, based on the information it gathered from sources. Blessed be the Gurman!

I see someone in the comments section of this liveblog has already asked the all-important question "Who's excited?" I think I am? Like I said, I just want to be able to relax after this. But more pointedly, is anyone excited for new Siri stuff? Do we want more changes to iOS? Tell us below!

Hello everyone and welcome to our liveblog of WWDC 2026! You're here a bit early, so you must be raring to get going. As am I, to be honest, but mostly because I want to get things over with! Let's get on with it! Here on the ground are myself and our senior reporter Karissa Bell, and we'll be sharing plenty of pictures of food, I assure you.