Google I/O 2026 LIVE: Android 17 latest, Gemini, Android XR and more

Setting expectations

Last week we wrote about Google I/O incoming – and the 3 biggest announcements we're expecting to see come out of the keynote.

"Gemini, Gemini, Gemini," could count for all three, but, no, there'll be more too. Android XR is due extra time, given the expectation of new Samsung Galaxy Glasses. Aluminium OS surely will get featured, too, given the recent reveal of Googlebook.

Android is still all about widgets

Android is 18 years old and widgets are still very much the focus. As Dieter Bohn announced at The Android Show last week, Gemini Intelligence will be able to create custom widgets.

I remember the launch of early Android phones. One groundbreaking device I reviewed at the time was the HTC Hero. That really made Android feel usable - and that phone was famous for its widgets.

Nearly two decades have passed, but simple widgets are still really useful.

Google I/O is 18 years old

While we're talking about history, let's talk about the history of Google I/O. The first I/O event was held in 2008. What does I/O mean? You'll think that it just means "input / output" from the I/O that's common in computing. Well, it's more complicated than that. I/O also means "innovation in the open", which is a bit of corporate marketing spiel - but the actual origins are a lot geekier.

Google has previously explained that I/O actually comes from the origins of Google's name - googol. "A googol is a one, followed by 100 zeroes, so that's where the I/O name and logo came from. We just took the first one and zero, and we left off all the other zeroes. It all goes back to Google," explained Lorin Platto, director of events at Google.

Are we going to hear more about Android 17?

Android 17 is well in development and even brands like Samsung are now running the Android 17 beta. While the software itself might not bring many changes - Material 3 Expressive was a huge redesign just recently - it's the features that might come with it that we're more interested in.

The Android Show has already talked about Gemini Intelligence and the huge change that's going to bring, while also confirming easier sharing between Android and iPhone, easier device switching and encryption in RCS messaging between platforms.

That means that Google has got some of the news out of the way - perhaps we'll get to hear when Android 17 will arrive for Pixel devices later today.

It's always good to take a break

I'm not suggesting you take a break from our Google I/O live coverage, but Google recently made an announcement to help you stop doomscrolling. It's called Pause Point.

This is an intervention from your phone to ask why you're opening a distracting app and give you time to reconsider, rather than just letting you doomscroll for hours. We're living in an attention economy where apps want to grab and hold your attention and this feature could be a useful intervention.

Let's recap on Android XR

Google, Samsung and Qualcomm were really public about their collaboration on Android XR, Google's latest attempt to crack face computing. You'll remember Google Glass skydived into Google I/O in 2012 as the company's first attempt at smart glasses, before Google Cardboard appeared in 2014 as a crack at affordable VR.

Now we're in the era of Galaxy XR running on the Android XR platform - with the expectation that we'll see real products at Google I/O in the smart glasses format. Meta has held the lead for too long with Ray-Ban and we're expecting to hear from Xreal with Project Aura and the partnerships with Gentle Monster and Warby Parker driven by Samsung. Android XR could be about to hit the mainstream.

It might be that Samsung's own brand Galaxy Glasses follow later in the year.

How is Google going to deliver agentic AI?

Agentic AI is the talk if the AI world. This is what takes the current question and answer chatbots that we all know - like Gemini - and makes them more useful as a personal assistant by giving them agentic powers. What is agentic AI? It's the ability of the AI to act as an "agent" on your behalf to get things done.

Currently Gemini can do things like create appointments from a list and add them to a calendar, but Google has broader ambitions for agentic AI. While Gemini Intelligence sounds great, we've also heard about something called Remy.

Think of Remy like a really, really, advanced version of Microsoft Clippy, using Gemini powers to become a proper personal AI assistant. Remy is all about getting things done for you, with task lists you can monitor to check progress. There's going to be a lot to consider around trust, privacy and control here.

What are you looking out for most during Google I/O?

There are plenty of rumours on what we can expect later – plus a few confirmations – but what are you hoping to find out more on the most?

We'd like to know your thoughts on the possible keynote announcements. Just choose the option below that most piques your interest. The answers will be revealed shortly before the event starts.

The Google I/O Keynote is just the start

We're going to be covering the Google I/O Keynote and tracking the biggest announcements right here, but that's just the start - a lot more detail emerges during the sessions of the actual I/O conference over the coming days.

That's where learn more about some of those big announcements, with sessions dedicated to Android, Chrome, Google AI and more. We're going to be follow Google I/O beyond the Keynote, to learn as much about what Google has in store for its users in the future.

Apple sends invites to WWDC on Google's big day, but these are different times

Hi, news editor Rik Henderson here for a while with additional pre-show thoughts and rumours.

First up, it's not exactly coincidental that Apple chose today to send out invites to its own developer conference starting on 8 June. Its keynote address will kick off at a similar time too – 10am PDT (6pm BST).

However, we might actually get a brief tease on at least one thing it's set to detail – the new Siri. That's because at the start of the year, Apple and Google penned a partnership for Gemini and Google servers to help bolster Apple's own AI services. Maybe there'll be a mention at I/O.

It might be time to upgrade your Android phone

We've seen a lot of rumours around Google's new systems and software, but one is more worrying than others.

It has been reported that some of the Gemini Intelligence features we'll learn more about during I/O might be restricted to certain handsets – those that are powerful enough to run them. And that could rule out the phone you own right now.

It is even being claimed that the more intensive capabilities will be locked to the most recent phones only, with some devices just a year old missing out. We'll hopefully find out more in just a few hours.

Google Home Display on its way?

As well as the likes of Android 17, Gemini Intelligence and Android XR, it has been rumoured recently that Google could debut a new "Google Home Display".

Potentially replacing the Nest Hub, the new smart device is said to be designed around Gemini (of course) and be the central controller for your entire smart home.

It's rare for Google to unveil new hardware during an I/O keynote, but we would be surprised if it wasn't mentioned, at least.

A veritable AI-fest

Let's face it, this year's Google I/O keynote is likely to be jam-packed with mentions of both "AI" and "Gemini". So much so, in fact, that you'd likely be rushed to hospital if you played a drinking game during it.

But rather than neck a tequila shot every time, as a bit of fun just let us know how many times you think Google will mention either "AI" or "Gemini" during its presentation on the slider below.

We're an hour out and excitement is building

The Google I/O 2025 Keynote kicks off at 10am PST / 6pm BST, so there's only an hour to go until showtime. So what's happened in the last hour? We're starting to see pictures from inside the auditorium as attendees start to file in. But online, we're also seeing people sharing the newly-updated Gemini app that some are getting access to. The reception seems positive so far.

What are we expecting from Google I/O?

Just recently I talked about the three announcements I was expecting from Google I/O. As we tick down to kick off - can we call it that? It's a World Cup year after all - here's what I'm expecting.

Gemini across everything. Not only a more powerful Gemini with a redesigned UI, but integrated everywhere and pulling everything together - Android, Android XR, desktop app, iOS, Googlebook, Android Auto and home devices.

New hardware is rare at Google I/O so I don't think we'll see the Pixel 11, but we've had hardware teased at I/O in the past, like Pixel Tablet. What we're all waiting for is a Google Googlebook, new Home hardware to bring Gemini in to rival Alexa, and advancements on the Nest front.

I think the hardware will appear with a detailed dive into Android XR: we know it's coming in 2026, it's a relatively new platform with big partners, so smart glasses without a screen would be my guess, with a tease of what can be done in Android XR with display glasses too.

AI tools will have some focus with the enhanced powers of Veo Omni already leaking out, but there's also Google Search and the advance in AI there, Google's Workspace apps and other Google properties.

But above all, I think we're going to be shown what Google continues to do with AI and the opportunities it represents for problem solving and productivity.

Just a few minutes to go

Hopefully you've taken a comfort break and refilled your giant Stanley cup to keep you hydrated through the Keynote, which you can watch on YouTube embed above. The Keynote is expected to last about 1 hour and 45 minutes.

Hopefully you also have your Google I/O Bingo card at the ready. I just asked Gemini what it should include, and shared it above.

And we're off with the Google I/O 2026 Keynote

As we get going it's Chris Hall on the live blog here, bringing you all the details as Sundar Pichai takes to the stage to open up the show.

We've just seen an inspirational video about Gemini, showing how Gemini can help everyone out in a huge range of different ways.

We open straight into talking about AI

Sundar Pichai opens the show, reminding us that it's 10 years since the company made the pivot to be AI first. The uptake of AI has been huge over the last year, with over 19 billion tokens processed each minute. It's mostly numbers at the moment.

Pichai tells us that AI Mode in Search is growing fast and that over 900 million people are using the Gemini app each month.

The first real announcement drops with Ask YouTube so you can use natural language to search YouTube, with roll-out expected in the US in summer.

Google wants to speed up document creation

We've been treated to a live demo though Gemini Live, showing that it's a much faster experience and making real-time changes to created Docs. It does look super fast, drawing from a number of documents to show off Gemini's multi-model skills, while removing the need to copy and paste the content into a Google Doc. You say what you want, you get a Doc at the end of it, perfect for drafting things.

Gemini Omni is official

Omni leaked recently as Google's next-gen video creation tool and Google says that it understand physics much better to create much better videos. Omni allows you use your own videos and edit them - using voice commands - the demo looks really powerful and easy.

It means you don't need all those graphical effects from editing software - Omni will be able to do it. Omni Flash is coming to the Gemini app straight away - with Omni Pro coming in the future.

Deepfake detection in the spotlight

Google is expanding its SynthID tool to detect AI created content. That will allow users to easily check if an image has been created or manipulated with AI. Content needs to be tagged, but Google has confirmed that Nvidia and OpenAI have both signed up, which should lead to greater transparency with images online.

If you've used Google Lens to examine images in the past, you'll be able to get an AI check in the same way.

Agentic AI comes to life in Gemini Spark

This is what we've been waiting for - Gemini Spark. This is a personal AI agent that will carry out tasks for you - probably what was called Remy in earlier leaks. It's powered by Gemini 3.5 and it runs on private Google Cloud servers in the background, so it will keep working even when you're not.

It's divided into tasks and allows you to manage what Spark is doing for you. Spark will carry out complex tasks, the example is planning a block party. It can create a spreadsheet to track replies, it can auto-update Docs and it can be tasked to do things like track the RSVPs and email reminders - while also compiling who is bringing what.

Spark works in desktop and mobile so everything is synced. Spark can act like a second brain - you can just talk to it, a bit like telling a personal assistant what you want. It's coming next week to Google AI Ultra subscribers first.

It will also come to Chrome to act as an agentic assistant directly from your browser.

Search expanding with new AI Search box

AI Mode is controversial for those who depend on traditional search results - and AI Mode is now getting an update. Use is doubling every quarter according to Google, probably because it's hard to avoid.

There's a new intelligent Search box, including AI suggestions for what you want to search for. You can search with images, files and voice, with the new Search experience rolling out today.

By combining with Gemini Spark, questions you ask can be agentic, so instead of just getting a snapshot at that moment, you can have updates in the future too. Basically, searching is now an AI function, it's not just a text input.

We haven't got there yet, but Samsung has confirmed Intelligent Eyewear

The Keynote hasn't reached Android XR yet, but Samsung has confirmed what's coming. You can find all the details right here if you can't wait.

Let's talk about shopping

Google has announced a universal shopping cart meaning you can add things to your cart from multiple places. AI will track the price history of things in your cart, while also using AI to suggest alternatives if it thinks you've picked incompatible products.

The universal cart can also suggest buying with different payment cards to make sure you get the best offers and once ready to check out, Gemini can do it all for you.

Google has confirmed a universal profile for different stores to power the new shopping experience. It looks like Google wants to own the online shopping experience to save you visiting all those different websites. Is that a good thing?

Gemini app gets loaded with Spark, Omni, Daily Brief and 3.5 Flash

Core to Google's AI experience for many is the Gemini app. Gemini 3.5 Flash is going to boost what Gemini can do, but before we get there, there's a new design to the Gemini app, called Neural Expressive, with a new Gemini Live experience - including regional dialects.

Gemini is going to be better at creating exciting layouts, including images and video so it's not just text. This will scare those with websites, as Gemini will basically create it all on demand.

Gemini Omni is also coming into Gemini as expected, meaning much more powerful video creation. It's also coming today for subscribers, so it's getting straight to work.

Gemini also gets something called Daily Brief. It's designed to be your first port of call in the morning so you know exactly what's going on. Add that to Gemini Spark and Gemini is suddenly a much more involved experience.

I think we're getting a glimpse at what your phone will be like once Gemini is the entire OS.

AI for creativity: Google Pics, Stitch, Tools, Music

Google has shown off new creative tools. Google Pics is a new image creation tool in Google Flow to allow image creation on the fly with AI - all marked with SynthID of course. Stitch will allow people to create webpages, meaning a faster route to getting that site live using AI - with simple voice inputs to make changes.

Google Flow gets Gemini Omni for advanced video creation, a new agent to add extra powers, more tools and music features. All this is coming to Google Flow for creatives.

Google's Android XR glasses are now official

We finally have confirmation (that we already knew) about the Android XR smart glasses. Calling them "audio glasses", it's official that they are coming from Gentle Monster and Warby Parker, with Samsung, and they will work with both Android and iOS.

We get a live demo of Gemini on Android XR, which is great for those who love using voice. Of course, the agentic skills will work here to do things for you - like place your coffee order without having to open the app and do it manually.

Anyone who has used Gemini on headphones might not have such faith in Gemini on glasses.

Google has also confirmed that glasses with a display will get Create my Widget functions for on-the-fly customisation - but those are coming sometime in the future.

That's all folks!

And that, as they say, is that. Let's just recap quickly: we've had talk about Gemini at every level, with new Gemini Spark agentic assistant that's going to be basically everywhere. We've seen Gemini getting a redesign, with more powerful video tools through Omni, and Live Doc creation in Gemini Live.

Google Search is also going agentic, with a new Search box that's also an AI tool, so we're well beyond just straight text searches now. Shopping gets super-charged with a universal cart and new protocols to help online stores all work with Google.

Finally, we're in the era of Android XR glasses, with the audio models, developed by Samsung with partners, now going live.

What we didn't get was more information about Google hardware, no Home or Nest products, no mention of future Pixels, nothing additional about Android 17 and not even a mention of Googlebook.

But, outside of the Keynote, we did get confirmation of new Gemini skills coming to Volvo and Xreal's Project Aura - so there's definitely more to catch-up on.