Google is doubling down on agentic AI to enhance Android customers’ on-device experiences by embedding Gemini deeper into the merchandise that individuals use every single day.
The tech big on Tuesday, Might 12, launched Gemini Intelligence, an AI agent-based working layer designed to learn a person’s display, transfer throughout apps, and autonomously full duties that might in any other case require customers to leap between a number of platforms.
The announcement was made at The Android Present 2026, a precursor to Google’s I/O developer convention subsequent week, the place the corporate previewed the most recent Android model, together with AI-powered app automation, smarter model of Chrome on Android, new instruments for creators, new safety features, and a revamped Android Auto expertise.
With Gemini Intelligence, Google is transferring past conventional chatbot-style assistants towards autonomous AI brokers. Whereas AI brokers usually are not new, their adoption remains to be within the nascent phases. Gemini Intelligence might doubtlessly mark a turning level for the mainstream adoption of AI brokers, particularly in India the place over 97 per cent of all smartphones are powered by Android.
Gemini Intelligence’s debut comes a month earlier than Apple is anticipated to unveil a revamped Apple Intelligence, reportedly powered by Gemini itself, at WWDC 2026. It highlights Google’s large benefit in client AI owing to its huge distribution community. Nevertheless, rivals comparable to OpenAI are additionally reportedly eyeing smartphones with AI brokers on the core.
“We’re reworking Android from an working system into an intelligence system. This implies deeply understanding the context, anticipating your wants, and getting issues achieved in your behalf,” Sameer Samat, the president of the Android Ecosystem at Google, mentioned through the keynote presentation on Tuesday.
How Gemini Intelligence will work
On the launch, Google revealed little about Gemini Intelligence underneath the hood, together with the way it was developed or which massive language fashions (LLMs) energy its agentic AI capabilities. Nevertheless, the tech big talked about that it spent months fine-tuning the multi-step job automation capabilities of Gemini Intelligence utilizing fashionable meals and rideshare apps on Galaxy S26 and Pixel 10 gadgets.
Story continues under this advert
Job automation permits Gemini to make use of sure apps on a person’s behalf. As an illustration, customers can ask Gemini to take a look at the visitor checklist for a barbecue, construct a menu, add substances to an Instacart checklist, and return for approval earlier than checkout. One other instance is asking Gemini to discover a person’s class syllabus in Gmail then placing the books of their cart.
Google mentioned job automation will probably be obtainable for a wider vary of apps “quickly”. It will likely be rolled out in phases, beginning with the most recent Samsung Galaxy and Google Pixel telephones this summer time, earlier than increasing throughout extra Android gadgets, together with watches, automobiles, glasses and laptops later this 12 months.
One other addition to Gemini underneath Gemini Intelligence is multimodality. As an alternative of manually switching between apps and duplicate pasting textual content, customers can, as an illustration, lengthy press the ability button on the grocery checklist saved on the Notes app, and Gemini will construct a procuring cart with all the objects for supply. Beforehand, Gemini might solely use voice or textual content prompts to take actions.
Different Gemini Intelligence options
Customized widgets: Customers can create customized widgets simply by describing what they wish to see on their display. As an illustration, customers can ask Gemini to generate a house display dashboard comprising weekly high-protein recipes for prepping meals, and cyclists can create a climate widget that solely tracks wind pace and rain.
Story continues under this advert
Customized Widgets utilizing Gemini Intelligence. (Screenshot: Google)
Autofill: Gemini Intelligence will pull related data from linked apps and mechanically fill out tiny textual content fields in kinds throughout Android apps, together with Google Chrome. Google emphasised that this function is opt-in and will be turned on or off in settings.
Autofill powered by Gemini Intelligence. (Screenshot: Google)
Rambler: This function converts speech into textual content messages whereas eradicating pauses and filler phrases like ‘uhms’ and ‘ahs’ to make messages extra concise and readable. Rambler will clearly present customers when it’s enabled, and the audio is just used to transcribe in real-time and isn’t saved or saved, as per Google. It additionally comes with assist for Hindi and different languages.
Rambler powered by Gemini Intelligence. (Screenshot: Google)
In addition to these options, Gemini Intelligence comes with an up to date design language that builds upon Materials 3 Expressive launched final 12 months. To deal with safety issues round AI brokers going rogue, Google mentioned that Gemini will at all times come again to the person earlier than finishing a transaction.
Who’s successful the AI distribution race?
Google is betting on Gemini to cement its lead in client AI, given its management over platforms and merchandise that individuals use every single day, which permits it to embed its AI instruments throughout the ecosystem at scale.
Story continues under this advert
Take Android Auto. On Tuesday, Google introduced a redesign of its in-car OS round Gemini, turning automobiles into one other key floor for its AI assistant. At the moment, Android Auto is in additional than 250 million automobiles. Google mentioned that the brand new model comes with its greatest maps replace in a decade and Gemini-powered assist with duties like ordering dinner whereas driving. Among the many first automakers to get a number of of the brand new Android Auto options embrace BMW, Ford, Hyundai, Kia, Genesis, Mahindra, Mercedes-Benz, Renault, Škoda, Tata, and Volvo.
Google additionally mentioned it’s bringing an AI looking agent to its fashionable Chrome internet browser on all Android gadgets. The AI looking agent can be utilized to analysis, summarise, and examine content material throughout internet pages in addition to auto browse and autonomously full duties comparable to appointment bookings.
In the meantime, Gemini Intelligence might add stress on Apple to show at WWDC 2026 (happening from June 8 to June 12) that its personal model of private AI on the cellphone is additional alongside. Notably, Google is a part of the iPhone-maker’s AI reset. Gemini fashions will energy Siri and some different Apple Intelligence options as a part of the multi-year deal between the 2 tech giants.
Apple’s iPhone is constructed round functions developed by third events, and this mannequin has performed an enormous function in making the iPhone what it’s in the present day. However it’s dealing with potential disruption not simply by Google but in addition by smaller gamers comparable to OpenAI. The Sam Altman-led startup is engaged on a mysterious new cellphone that might closely depend on AI brokers, based on a analysis notice from trade analyst Ming-Chi Kuo final month.