A looming Android reconstruct is set to change the way smartphones work for millions of users worldwide. Ahead of their annual I/O developer conference, Google has detailed a vision for Gemini’s deeper utilitarian integration across all apps, among other improvements. It essentially transforms Android phones into artificial intelligence (AI) devices, even as industry whispers suggest a separate new category of these devices remains somewhere over the horizon.
Core to a modernised Android is a new Gemini Intelligence layer, which evolves Gemini from a conversational chatbot into an active, contextual assistant across third-party apps on a user’s Android phone. AI’s intent to automate tasks where possible — a significant change in how you’ve used an Android phone.
“We believe that an intelligent system, and an agent that it empowers, should just get things done for you, especially the more mundane, boring daily tasks. Another thing is that it should meet you where you are and the apps that you’re already in with proactive suggestions,” said Dieter Bohn, director of product operations at Google.
Some of Gemini Intelligence capabilities include bringing context to auto-fill when filling forms across websites and apps, a Rambler feature for the keyboard app which can convert typically spoken sentences with fillers into cogent messages or lists, and automating certain tasks such as booking tickets for an event you’ve discovered.
Google is working with developers to ensure they have the tools for the Gemini Intelligence integration. Clear guidelines, and tighter integration within Android (and its subsequent updates) will minimise the chances of feature and compatibility fragmentation later.
Ben Greenwood, director and product manager for Android Core Experiences, noted that these new features will roll out in waves through the coming months, starting with flagship Android devices, before covering more phones. Android’s current flagship crop includes Google’s own Pixel 10 phones, Samsung’s Galaxy S26 Ultra as well as the Xiaomi 17 Ultra and Vivo X300 series.
When asked by HT about a possible subscription layer defining usage and utility of Gemini Intelligence as core to the Android experience on phones, Greenwood confirmed that base features will remain broadly available to all devices, but with a caveat. “There are some usage limits for higher usage, and in those cases, you can upgrade to an AI subscription plan,” he said.
Limits on the free tier and how those stack up to paid plans will become clearer when these features are accessible at an OS level, a scenario where usage is expected to be more. In India, Google has three paid AI plans — AI Plus that costs ₹399 per month, AI Pro for ₹1,950 per month and AI Ultra priced at ₹24,500 per month.
“Starting in late June with Chrome Auto browse, Gemini inside Chrome will be able to help you with everything from appointment, party planning, or searching for an out of stock item, from your Android phone,” Greenwood detailed the first step of an elaborate plan.
Google is making a strong move ahead of Apple’s annual WWDC keynote next month, where a revamped Apple Intelligence suite is expected to be announced. Earlier this year, Apple and Google announced a partnership to use Gemini AI models, including for a conversational Siri, expected later in 2026.
On the privacy question, Google insists its philosophy for Gemini Intelligence is a mix of on-device processing and the cloud. For instance, Rambler will process voice transcriptions on the cloud (which means it’ll require an active 5G, 4G or Wi-Fi) but since there’s no storage or memory, its returned back to the device after processing.
“Some are delivered on device, while published and verifiable private technologies in the cloud offer similar privacy behaviours. These are all investments we’re making to ensure that across intelligence features, users can have the best quality experience, and be safe and private,” Greenwood said.
Google had introduced Private AI Compute in November, allowing Gemini models to securely process user data within a specialised, protected space.
Data that Gemini Intelligence gets access to, such as for filling web forms and attaching documents, depends on what users choose to share with Google. This could include saving a scan of your passport or driving licence in Google Drive or Gmail or Google Photos.
It remains to be seem whether there is a possibility to opt out of certain data sharing, specific features or the entirety of Gemini Intelligence on Android devices.
“It’s about making technology that truly understands and works for you, getting your intention and translating it straight into action, and it’s woven in so it’s there when you need it, and it fades into the background when you don’t,” Greenwood added, noting this is the first wave of Gemini Intelligence.
Warm handshake
Among other key updates to Android is something Bohn called “tools to disconnect when you need to”. A problem Google is trying to solve, is one of a human tendency towards unbalanced attention and distraction.
“We’re all guilty of popping into our phones to do something quick, and then you end up on autopilot and in some app, and suddenly an hour has gone by. App times are one solution for this, but they’re easy to ignore without thinking about it,” he noted.
There will be pause points for apps that you select as ‘distracting’, and when that app is opened, a prompt screen will instead suggest a number of possible activities such as breathing exercises for a few seconds, or swiping through a collection of photos first.
The idea is for the mind to reconsider the choice of opening this particular app at this time of the day. At the bottom of the screen will be an open clearly labelled “don’t open”. The intention is to build, or in some cases, repair a habit of usage.
Bohn insisted there are more digital well-being tools in the works, which will be coming later this year.
Google wants to reduce friction in movement between devices. First, new switch device mechanism will now also transfer saved passwords as well as the Home Screen layout, along with contacts, photos, messages and apps, when a user switches from an existing Android device to a newer one.
The updated device transfer methodology will roll out later this year for Samsung’s Galaxy phones and Google Pixel phones, with wider support in the months after that.
There is also a concerted effort to make Android phones and Apple iPhones handshake better in terms of being able to send and receive files wirelessly. It was late last year when Google made Quick Share compatible with Apple’s AirDrop standard, for users who’d like to send something from their Android phone to a friend’s iPhone, or vice versa.
HT noted that some phones have integrated that, while some haven’t — Samsung Galaxy S26 Ultra and the latest Vivo X300 Ultra can Quick Share to an iPhone, but some phones such as the Xiaomi 17 Ultra still cannot. Google says the reason is likely missing supporting hardware, but expect parity with the next generation of phones.
