Google unveils Gemini Spark, a 24/7 AI agent with deep Gmail and Workspace integration

Google introduces Gemini Spark, a 24/7 agentic assistant with Gmail integration | TechCrunch

In the race to build compelling personal AI agents, Google may have an underrated advantage: it already has all your emails. At the company's I/O developer conference on Tuesday, Google announced a new agentic personal assistant called Gemini Spark that was built from Gemini base models and an agentic harness from Google Antigravity. Alphabet CEO Sundar Pichai described Spark as the next evolution of smart digital assistants, using agentic AI to take on long-horizon tasks with minimal oversight. "It's your personal AI agent that helps you navigate your digital life, taking action on your behalf and under your direction," Pichai told reporters during a pre-briefing of the product. "It runs on dedicated virtual machines on Google Cloud seamlessly, [so] you don't need to keep your laptop open to make sure it's running." Spark follows a wave of popular agentic products from major AI labs, most notably Anthropic's Claude Cowork and OpenAI's ChatGPT Agent, but it will have particular value because of its integration with the larger Google suite of products. Spark will include out-of-the-box integrations with Gmail, Google Docs, and other Google Workspace products, saving users the work of setting up connections and permissions with outside apps. Users can email Spark directly through a dedicated Gmail address, and the agent can interact with the web directly through Chrome. On mobile, you'll also be able track the agent's progress through the new Android Halo system. "Need to send an email to your boss with a status update? Spark can pull all the facts from your emails, your docs, your sheets, and slides and write the draft for you," said Google Labs' Josh Woodward, VP of the Gemini App and AI Studio. "Small businesses are using Spark. They can watch over their inbox, so they never miss a question from a customer." Like other agentic assistants, Spark can be integrated into a wide range of services over MCP, and Google expects to roll out more connections in the months to come. Spark is currently in testing at Google; the company expects to make it available to Google AI Ultra subscribers next week.

Google's Response to OpenClaw's 24/7 AI Agent

Gemini Spark is Google's take on a steroided-out assistant agent that knows everything about you, announced as part of the company's updates to its Gemini chatbot app at this year's I/O developer conference. Software companies have been talking up AI agents for some time now, but I wasn't impressed until I tried Anthropic's Claude Cowork in January. I sat back as the bot organized the scattered screenshots littering my desktop into labeled folders without a single click, and felt convinced that this might be a turning point for how people interact with their computers. Many other early adopters in San Francisco experienced similar moments when they set up the mega-viral OpenClaw bot earlier this year, not just to help complete a few tasks but to run their whole online lives. Power users attempted to fully automate their inboxes, calendars, and text messages, and even run a vending machine to varying levels of success via OpenClaw. It's not without risks -- you have to give these agents control of your data and computer, and OpenClaw almost deleted an entire trove of emails for one Meta employee who was experimenting with it Whether it's my daily schedule via Google Calendar or my date-night dinner spots through Gmail confirmations, Gemini Spark can dive deep into the well of my personal info before I even connect to a third-party integration. While the standard Gemini app can complete many of the same tasks, Sparks' differentiator is that it proactively gathers details and takes action while you're away, rather than waiting for you to prompt it. Google pitches Gemini Spark as a one-stop shop for completing tasks people previously handled manually or did in other apps. The agent can look through your credit card bill regularly to flag surprise fees -- sorry, RocketMoney app, won't be needing you anymore. Spark can be calibrated to automatically skim every email about your preschooler and highlight key dates for a morning digest report. You can even throw all your meeting notes at Spark and ask it to draft a Google Doc and generate follow-up emails to the right people. This agent is getting a slow rollout, arriving for a small group of early testers this week and launching next week in beta for subscribers to Google's $100+ per month AI plan. It's pricey to be one of the first people to experiment with Spark! The company plans to allow Spark to connect through Gemini to third-party apps, like OpenTable and Instacart, for additional automation opportunities in the coming weeks. Other features imminent on the Spark road map include allowing the agent to manipulate your local browser and the ability to text or email commands to the agent. Being able to text commands to your agent sounds like a key factor in actually making the Spark experience feel seamless. Rather than opening the Gemini app and getting distracted, I'll spend all day texting Spark my increasingly niche requests, as if it were assistant Andrea from The Devil Wears Prada. One of the main measures of success when trying this agent will be how often it goes off the rails. "Spark operates under your direction," reads Google's announcement blog about the agent. "You choose whether to turn it on and what apps it connects to, and it's designed to ask you first before performing high-stakes actions like spending money or sending emails." Anyone who tries the tool is taking a risk by using experimental software that's powered by personal data. Google plans to expand the agentic shopping feature to allow users to set spending limits and preferred merchants that Spark will adhere to, though exercising caution is critical. "We think of it as if you're giving a teenager their first debit card," says Josh Woodward, vice president of Google Labs and the head of the Gemini app. Much like the changes Google is implementing in Search, which brings agentic task automation without needing to leave the search experience, Spark is Google's chance to push AI agents further into the public zeitgeist. Let's see if it has the necessary spark to pull it off.

Gemini Spark Gives Google Way Too Much Access to Your Data

Dashia is the consumer insights editor for CNET. She specializes in data-driven analysis and news at the intersection of tech, personal finance and consumer sentiment. Dashia investigates economic shifts and everyday challenges to help readers make well-informed decisions, and she covers a range of topics, including technology, security, energy and money. Dashia graduated from the University of South Carolina with a bachelor's degree in journalism. She loves baking, teaching spinning and spending time with her family. Google's shoving more AI into its software and devices. Meet Gemini Spark, the new AI agent that's like a 24/7 personal assistant. Google debuted the tool on Tuesday at its Google I/O developer conference. We saw a demo of Gemini Spark in action, handling multistep tasks, like planning a block party. Gemini Spark counted RSVPs, followed up with those who hadn't and created an RSVP tracker that automatically updates when new responses come in via Gmail. Gemini Spark will be available to some testers this week and available in Beta for Google AI Ultra subscribers next week. There are also plans for Gemini Spark to be available as a Mac desktop app and, later, on Chrome for everyone. The new AI agent sounds like a dream if you're busy or overwhelmed by the thought of umpteen tasks on your to-do list. But ask yourself if you're really OK with Google having access to all that information. The big question is, how can Google guarantee your personal data is safe and secure when giving Gemini Spark access to apps including (but not limited to) Gmail, Calendar, Drive, Docs, Sheets and Google Maps? Google says these connections are turned off by default, but you can turn them on. And Google says Gemini Spark "does not read your emails indiscriminately." But when these connections are enabled, we still don't know which information is being stored and shared for Gemini Spark to work. And while we hope that Gemini Spark sticks to only Google apps, what about other information stored on your device, like the guests' phone numbers if they RSVP'd to that block party or their emails used for follow-ups? Despite the advanced security measures most AI tools promise, it's only a matter of time before scams and data breaches occur. Google says that Gemini Spark works in the background using Gemini 3.5 Flash and Antigravity. It will run even if your laptop and phone are turned off. Gemini Spark will run independently, but under your direction. If it's running in the background on its own, I'm concerned it's watching my screen like a hawk to complete tasks independently. This makes me question how Google can guarantee my data is safe. Giving an AI agent permission to scan my Google Sheets and highlight important information may save me time, but imagine that data falling into the wrong hands for work or personal matters. That's always a risk, but Gemini Spark feels like an invitation. For instance, letting Gemini Spark order snacks for my son's soccer game and having Instacart deliver them gives room for errors beyond too many oranges and juice boxes. Think about your payment information and addresses being used in the process while you sleep. There are too many risks to my personal data and devices. I'll pass.

Google is launching its own version of OpenClaw

Gemini Spark is powered by the newly introduced Gemini 3.5 Flash and runs in the background 24/7 using virtual machines on Google Cloud. The AI agent will connect to Workspace apps like Gmail, Docs, Sheets, and Slides, but Google is expanding integrations to third-party apps using the Model Context Protocol (MCP), an open standard that allows AI models to plug into external systems or data. That includes Canva, OpenTable, and Instacart. This summer, Google will give Spark the ability to interact with your local files through the Gemini app on macOS. "Even when you close your laptop or turn off your phone, Spark can keep working in the background as you go through your day," Josh Woodward, the vice president of Google Labs, Gemini, and AI Studio, said during a briefing. "When you use it, it almost feels like you're tossing things over your shoulder, Spark's catching them, and gets the job done."

Is Spark Another Rebrand of Google Gemini or Something Else?

MOUNTAIN VIEW -- I wouldn't blame you if you came away from the recent Google I/O keynote with no idea what Gemini Spark is. After all, Gemini began its artificial life as Bard, Google's first-gen AI chatbot. Based on what we saw during the keynote, it might seem that Gemini is now transmogrifying into Spark. But that's not correct. Unlike a traditional AI chatbot, where you input a prompt and wait for a response, Gemini Spark, the company's new proactive AI, lets you set it and forget it -- kind of. It runs in the cloud, so the process continues across your Google account, in a Workspace suite, and, eventually, third-party apps, depending on what you ask it to do. Unfortunately, Gemini Spark's debut during the keynote only added a confusing storyline to Google's AI platform, rather than generating excitement about what's to come. There are many elements to Gemini, the product. But because of Google's overt insistence on claiming it's one of the first to deliver on the promise of consumer-ready, proactive AI, it's not immediately obvious that Spark is the manifestation of that. I attended a few Gemini demos at the developer conference to better understand what's going on. If anything, Google wants you to see Spark as your AI partner. Gemini Spark: The Active Partner The first demonstration I hit up at the event showed me how Gemini's existing AI Mode integrates a proactive feature as you query the search engine. The initial query asks Gemini to compile a list of upcoming concerts in a specific music genre and region. This seems like something you would do with the new Gemini Spark, too, and then have it update over time. However, the functionality I saw is contained entirely within AI Mode, just on search. But then I started putting two and two together and realized that this is essentially a fork of what Google told us earlier about what Spark could do. Hours later, after trudging through the sun-soaked grounds of the Shoreline Amphitheater, I found myself in an actual Gemini Spark demo in an air-conditioned room. There, it started to make more sense. Spark is the assistant, the agent that Google wants you to tap into when you need some long-form work done. For instance, Google's main examples on-site were wedding preparations and home renovation projects; you can use Gemini Spark to help you draft everything from the initial emails you send to vendors and contractors to listing the price differences and negotiations, then keep it updated as things progress. Gemini will look through your email (and other Google Workspace apps you set up) to continually populate that data, though you can also feed it information as needed. These are things you would have done yourself over time anyway, but Spark does them automatically. Like the AI mode concert example I saw earlier, it's a proactive agent that fetches what you need. This is the exact agentic trajectory we expected from Google in the first place. We were promised a future where, even without coding experience or a deep knowledge of the technology landscape, you could type a prompt in casual, everyday language and ask an AI to help you get your life situated. That kind of practicality is what will ultimately set Gemini Spark apart from its competition among the general populace. The official Google Spark beta launches soon, with the full capability going live this summer for paying Gemini users. I'm looking forward to integrating this feature and using it to help me manage my household, where I currently serve as the sole, exhausted "project manager." With Spark deeply integrated across the Google and Android ecosystems, I won't have to fumble over my words copying a note into Google Keep or fire off an email to the right address. If Google can get out of its own way and fix the branding of it all so that it makes more sense, Spark might actually deliver on that elusive promise of an AI that actually works for you even when the screens are off.

Google's Spark Uses Gemini AI to Help Plan Your Life

Expertise Smartphones | Gaming | Telecom industry | Mobile semiconductors | Mobile gaming At Google I/O 2026, new AI models and features are spilling off the presentation stage. One, called Gemini Spark, is what Google describes as a personal AI agent that runs 24/7 taking actions on your behalf. As described on stage, Spark's purpose is a little broad, but it's meant to run projects in the background, collecting data from the web and your various Google software accounts (like Gmail, Docs, chats and others). Google's presenter provided a few examples of how Spark can help. One was sending a work email to the team, compiling info from documents and messages into a summary of wins from the week. Another, bigger task was planning a block party, tallying RSVPs, keeping track of who was bringing what and reaching out to folks who hadn't responded. Then, Spark can generate files to show progress -- like an RSVP tracker that updates whenever it detects email responses in Gmail. The pitch for Spark is to account for all the little things that would've otherwise slipped your mind. In the party-planning example, it could identify some limits and considerations to avoid breaking your homeowners' association rules. And since it's humming in the background, it'll update as new information comes in. In another scenario described later in the presentation, Spark could see on your calendar that you're assigned to bring snacks for your kid's tee-ball game and order food on Instacart to arrive on time. The presenter pitched more assistant-focused uses for Spark, like listening to your brain dump of thoughts and organizing them into shape. Examples given on stage included reminders to reach out to a new friend from a party the previous night, or asking the AI agent to make a list of everything parents need to take care of for their student kid before summer, listed by deadline. This aligns with other AI features shown off at I/O 2026, like Docs Live, which similarly organizes brain-dump ramblings into coherent text. Spark will be coming to trusted testers this week and Google AI subscribers next week, but only for those signing up for the Ultra tier -- either the new $100-per-month level or the newly reduced $200-per-month level. For everyone else, Spark will be available in Chrome later this summer, allowing you to use it directly in the browser. Later in the year, Google will release Android Halo, a dedicated home base for AI agents on phones, and presumably Spark will be included.

Google reveals it's big plan to take on OpenClaw as the race for agentic AI begins

* Google unveils Spark, an agentic AI that runs tasks for you 24/7 on Google Cloud * Currently connects to Gmail and Docs, analyzes uploads overnight; macOS desktop actions planned * Rolling out to testers and US AI Ultra subs soon, with wider release likely When OpenAI brought LLMs into the greater public eye, big companies scrambled to get their own chat bots up and running. We saw big competitors like Copilot, Gemini, Grok, and Claude take to the scene, while smaller AI models (remember Adobe's one?) popped up here and there. Then OpenClaw made its debut, and suddenly, chat bots were old news. Now, it's all about agentic AI, allowing models to perform tasks in the stead of the user. In response, we've seen companies reveal their own agentic models, and Google just announced its own version: Spark. Google's Gemini Intelligence can fill your shopping cart straight from your notes app Agentic mobile AI is here. Posts By Simon Batt Google Spark wants to help you get more done without you lifting a finger Chat bots are so last month In the company's annual Google I/O presentation, the company pulled back the curtain on Spark. This is the company's big push for agentic AI, and its main focus is allowing Spark to ruin 24/7 using Google's cloud servers. That means the AI can chew through those laborious tasks while you sleep. Right now, Spark is a little limited on its capabilities. It can connect to your online services like Gmail and Google Docs, and analyse documents you give it overnight. However, Google does say it wants to bring desktop actions to Spark on macOS this Summer, which will make it more akin to OpenClaw and Claude Code. Deals Score Software & AI Subscription Deals for Smarter Workflows Unlock discounts on AI software and subscriptions -- explore offers for cloud credits, productivity suites, automation tools, and developer platforms. Save on trials and upgrades to add agentic features, integrations, and AI-powered workflows. Deals Explore Software, AI & Subscriptions Deals Right now, Spark is arriving to testers, and will appear for Google AI Ultra subscribers in the US in the coming week. However, it does seem that Google wants to roll this out as far as it can, so it likely won't be an Ultra exclusive perk for long. Granting OpenClaw access to your Google apps just got a lot easier Handle the app with care, though. Posts By Simon Batt

Google's Gemini Spark is an agentic AI assistant - Engadget

Google has announced a "24/7 personal AI agent" called Gemini Spark at this year's I/O developer conference. The company says Spark transforms Gemini from a standard AI assistant to an "active partner" that actually perform tasks for you. Spark is powered by Gemini 3.5 and is deeply integrated with Google Workspace apps, including Gmail, Docs and Slides. You can teach it to perform various tasks, such as creating a list of critical deadlines in your Gmail and sending it to you, or writing up a summary of ongoing updates in lengthy email threads. It's also possible to set it do to recurring tasks, like making it spot hidden fees in credit card bills every month. You can program it to complete several interconnected tasks for complete workflows, as well. For instance, you can ask Spark to look at meeting notes in your chats and emails and create polished reports in Google Docs, as well as draft an email that you can send along with that report. Gemini Spark is rolling out to testers shortly before making its way to Google AI Ultra beta users in the US next week. Spark is completely opt in: It's up to you whether to turn it on and use it, and you can also choose which apps it can connect to. In addition to Workspace apps, it will be able to link to Canva, OpenTable and Instacart right now, with more partner apps coming in the following weeks. Google will also release more features for the AI agent in the coming weeks, including being able to give it the ability to send texts and emails and the ability to operate your browser. The company assures that Spark will ask you first before performing "high-stakes actions like spending money or sending emails." Google also says that it's bringing Spark to the Gemini desktop app this summer so that it can access files and perform tasks on your computer.

Google takes on OpenClaw with Gemini Spark, a 24/7 AI agent for your digital life

Serving tech enthusiasts for over 25 years. TechSpot means tech analysis and advice you can trust. What just happened? It's been impossible to avoid the impact that OpenClaw has had on the AI industry. Google knows all too well how popular the agentic personal assistant is proving, so it's announced a rival: Gemini Spark. Alphabet CEO Sundar Pichai revealed Spark at the Google I/O developer conference yesterday. He described it as "your personal AI agent in Gemini app that helps you navigate your digital life, taking action on your behalf and under your direction." Google says Spark runs on dedicated virtual machines in Google Cloud, allowing it to keep working in the background 24/7 even after a laptop is closed or a phone is locked. It's powered by Gemini 3.5 Flash and the Google Antigravity harness, the same agentic framework Google is pushing into developer tools. OpenClaw's appeal has been its ability to work across messaging apps and personal workflows, handling email, calendars, files, and other chores like a digital worker that never clocks off. Spark is Google's attempt to package the same idea inside Gemini and Workspace, but with the obvious advantage of deep native integration with services millions of people already use every day, including Gmail, Calendar, Drive, Docs, Sheets, Slides, YouTube, and Maps. Third-party integrations through MCP are also coming, starting with Canva, OpenTable, and Instacart. Google gave examples of what Spark can do. The agentic is able to watch an inbox, summarize the most important weekly updates, create a prioritized to-do list, schedule calendar blocks, organize Drive files into a spreadsheet, or extract details from emails and log them into Sheets. Spark can also be taught "skills," such as reading a user's last 50 emails to create a personalized writing guide for future drafts. Google Labs VP Josh Woodward said during I/O that Spark can pull facts from emails, docs, sheets, and slides to write a status-update draft for a user's boss. He added that small businesses are using Spark to monitor inboxes so they don't miss customer questions. Woodward also offered a description of how Google envisions Spark: "When you use it, it almost feels like you're tossing things over your shoulder." There are obvious privacy and safety questions when an AI agent can live inside your inbox and documents. Google says Spark's app connections are off by default, users choose what it can access, and it will ask before carrying out actions such as spending money or sending emails. At last year's I/O, Google showed Gemini smart replies that could mimic a user's writing style by analyzing previous emails and Drive files. Spark pushes that concept much further, from drafting suggested replies to letting an agent act across a user's digital life. Gemini Spark is rolling out to trusted testers this week. A beta for US Google AI Ultra subscribers is scheduled for next week, with more features due over the summer, including the ability to email or text Spark, create custom sub-agents, use Chrome as an agentic browser, and operate through the Gemini app on macOS.

Google announces Gemini Spark to quietly run your digital life for you

Google says Spark will eventually be able to create custom sub-agents, shop on your behalf, and have deeper integrations with services like OpenTable, Instacart, and Chrome. Google just revealed a major redesign for the Gemini app as well as ambitious new features that take the chatbot well beyond its current capabilities. One of those new features is a powerful new AI agent called Gemini Spark. Think of Gemini Spark like a real-life personal assistant that quietly and intelligently works in the background to handle digital tasks on your behalf. At least, that's the vision Google is presenting and hoping to deliver. Spark is designed to stay active in the cloud even after you close the Gemini app, lock your phone, or shut your laptop. Instead of waiting for commands one at a time, Spark can proactively manage workflows for you across different services and connected apps. Google says the more you use Spark, the better it gets at learning your personal preferences and executing tasks on your behalf. Google shared several examples of how this would work for regular users, including Spark's ability to scan credit card statements to spot forgotten subscriptions, keep track of school emails and deadlines, organize messy meeting notes into polished documents, and even generate and send emails on your behalf. The company says Spark is powered by its new Gemini 3.5 Flash model and deeply integrates with Workspace apps like Gmail, Docs, and Slides. Eventually, it will also be able to connect to Google Chrome. Google is also preparing Spark for more advanced tasks. In the coming months, the AI agent will reportedly be able to make purchases on your behalf using Google's new Agent Payments Protocol (AP2), though the company says users will still need to approve high-stakes actions. Going forward, Spark is also expected to connect with third-party services like Canva, OpenTable, and Instacart. Google even plans to let users text or email Spark directly in the future. A version of Spark will also be available to Google's enterprise users. In Gemini Enterprise, Spark will be able to automate recurring tasks and execute multi-step work for users. It'll also support existing Gemini Enterprise connectors, including Microsoft SharePoint, OneDrive, ServiceNow, and many others. Google says Spark will soon roll out in the Gemini Enterprise app. Moreover, Google also revealed that Gemini Spark is heading to macOS in a much bigger way later this summer. Spark integration inside the Gemini desktop app on macOS will allow the AI agent to work with local files and automate workflows directly across the desktop.

Google Comes for OpenClaw With Gemini Spark, a Personal AI Agent

The open-source OpenClaw launched the start of the personalized AI agent craze, and now the major players are getting in on the action with their own walled-garden offerings. Enter Gemini Spark, Google's "personal AI agent" offering announced during the company's Google I/O developer conference on Tuesday. According to Google, Spark is designed to be a “personal AI agent that helps navigate your digital life." It will ostensibly allow users to hand out more complicated, multi-step tasks and trust Gemini Spark to handle them autonomously in the background across multiple platforms. In a demonstration of the agent's capabilities, Google's Vice President of Google Labs, Gemini, and AI Studio Josh Woodward showed Spark working to draft an email on his behalf. Seems simple enough. But Woodward tasked the agent with collecting information from Google Docs, emails, and chat conversations before running a skill designed to draft the email in Woodward's voice. In a second demo, Woodward showed Spark compiling a list of people who RSVP'd to a block party and building a document that itemizes what each person is planning to bring. That document auto-updates when a new email related to the event comes into Woodward's inbox. Google offered some other examples of what Spark can do via a blog post, including setting recurring tasks or triggers, learning new skills like sending a digest of events to relevant contacts, and creating workflows from raw material like meeting notes. It's not too difficult to imagine the functionality of Spark, though it does seem quite dependent on being deeply ingrained in the Google ecosystem. The company did announce that it will add functionality across Gemini-connected apps like Canva, OpenTable, and Instacart, which will allow Spark to work across those platforms. It's also promising the ability to text and email with Spark and create sub-agents to handle complex tasks. Per Google, Spark will be powered by the company's Gemini 3.5 Flash model, the latest iteration of its flagship model that it also announced on stage at Google I/O. It'll run on dedicated virtual machines that are set to be available 24/7, meaning you can give Spark a task and let it run in the background until it's completeâ€"and it'll keep working even if you close your laptop. Spark will reportedly be available on Android, iOS, and via web app to start, with plans to allow it to operate directly in Chrome in the near future. It'll open to "trusted testers" this week and get a wider release as a "beta" product starting next week, with Google AI Ultra subscribers getting access. Google announced that its high-end Ultra plan price will drop to $200 per month, and it's also introducing a new $100 lower-end Ultra tier (maybe just call it something else, eh?) that will have access to Spark.

Google launches Gemini Spark agentic AI assistant at I/O 2026

Google announced Gemini Spark at I/O 2026, an always-on agentic AI assistant built on Gemini 3.5 and the Antigravity harness. It runs on dedicated cloud VMs, integrates natively with Gmail and Workspace, and will be available to AI Ultra subscribers next week.[/tldr] Google has unveiled Gemini Spark, an always-on personal AI agent that can receive tasks via a dedicated Gmail address, browse the web through Chrome, and work around the clock without requiring users to keep a laptop open. The company announced the product at its I/O 2026 developer conference on Monday, positioning it as its most ambitious move yet in the intensifying race to build autonomous digital assistants. Spark is built on Gemini 3.5 and powered by the Antigravity agent harness, Google's newly expanded platform for building agentic software. It runs on dedicated virtual machines on Google Cloud, which means it can execute long-running tasks in the background without tying up a user's device. CEO Sundar Pichai described it as a "personal AI agent that helps you navigate your digital life." The product's most distinctive feature is its tight integration with Google's own ecosystem. Users can email Spark directly through a dedicated Gmail address, much as they would message a human colleague, and the agent can pull context from Gmail, Google Docs, and other Workspace applications without requiring manual setup. That out-of-the-box connectivity gives Google a structural advantage over rivals whose agents must rely on third-party integrations to access the same services. Josh Woodward, VP of the Gemini app, Google Labs, and AI Studio, has been leading the product's development. Under his stewardship, the Gemini app now serves more than 900 million users across 230 countries. Spark also supports the Model Context Protocol (MCP), an open standard that lets it connect to a wide range of external services beyond Google's own suite. On mobile, users will be able to monitor what their agent is doing through Android Halo, a new notification layer that surfaces live status updates at the top of the phone screen. Google says Halo will arrive later this year with Android 17, effectively turning the operating system into a dashboard for persistent AI agents. The launch follows a wave of competing agentic products. OpenAI recently merged ChatGPT and Codex into a single agentic platform under co-founder Greg Brockman, while Anthropic's Claude Cowork lets users hand off tasks that the AI then completes by controlling a desktop computer. Salesforce has also joined the fray, transforming Slackbot into an agentic system with more than 30 new AI capabilities. Where Google believes it can differentiate is depth of integration. Because it controls the operating system, the browser, the email client, and the cloud infrastructure, Spark can operate across all of those layers without the friction that standalone agents face when stitching together disparate tools. Whether that advantage proves decisive will depend on how well Spark performs once it leaves Google's internal testing environment. Gemini Spark is currently being tested internally at Google. The company plans to roll it out as a beta to Google AI Ultra subscribers in the United States next week. The AI Ultra tier, which Google also announced it is cutting from $250 to $100 per month, includes five times the usage limits of the existing AI Pro plan along with 20 terabytes of cloud storage and YouTube Premium.

Google I/O 2026: Gemini Spark and Gemini Voice Coming to MacOS This Summer

Speaking from Google's Mountain View, California, HQ during Google I/O, Josh Woodward, VP of Google Labs, Gemini app and AI Studio, demonstrated how you can select a bunch of documents you've uploaded onto your MacBook -- in his example, he used paperwork related to his pets, like vaccination records and allergy lists -- and then long-press the function key and verbally dictate what you want the AI assistant to do with it. The demonstration showed you can not only dictate an email relating to the documents (as well as requesting that it sound "friendly"), but at the same time also request that Gemini turn the files you selected into a table. You then release the function key, and Gemini will put together your multiple verbal requests. "What it's done is because I've selected those files in Finder using its multimodal understanding, it can go through the PDF, it can go through these images of their invoices, and it's all controlled by my voice, so it can actually take all that complex information and ... it's got a table in line," Woodward said. You can download the Gemini app for MacOS at gemini.google/mac, and then use a simple shortcut to launch it. If you've got a MacBook, you can access Gemini at any time by pressing Option and Space on the keyboard. It already includes some of Gemini's best features, like Nano Banana image generation. While most people use AI chatbots via their phones or web apps, a native app for a desktop operating system is a newer development. With Gemini poised to power Apple's new AI-redesigned Siri, it's unsurprising that MacBooks would also get Gemini built in. Gemini's voice capabilities and Gemini Spark will be arriving on MacOS in the summer.

Google's new Spark agent works 24/7, even when your phone is locked

Karandeep Singh Oberoi is a Durham College Journalism and Mass Media graduate who joined the Android Police team in April 2024, after serving as a full-time News Writer at Canadian publication MobileSyrup. Prior to joining Android Police, Oberoi worked on feature stories, reviews, evergreen articles, and focused on 'how-to' resources. Additionally, he informed readers about the latest deals and discounts with quick hit pieces and buyer's guides for all occasions. Oberoi lives in Toronto, Canada. When not working on a new story, he likes to hit the gym, play soccer (although he keeps calling it football for some reason🤔) and try out new restaurants in the Greater Toronto Area. Roughly four days ago, we reported on early leaks that pointed at Google developing a Claude Cowork-like semi-autonomous agent. Today, at Google I/O 2026, Google has officially pulled back the curtain on Spark. Related Gemini Spark screenshots reveal an AI that actually gets things done on its own Google's new AI agent can handle your email, your calendar, and way more without asking for permission Posts By Mark Jansen Described as a "24/7 personal AI agent" that "helps you navigate your digital life," Spark marks a major shift for Gemini, with the AI agent going from a tool that can answer your questions to an active partner that does real work on your behalf and under your direction. During its initial run, Spark will already be integrated with Workspace tools like Gmail, Docs, Slides and more. Powered by Gemini 3.5, the proactive tool stays active in the cloud even after you close the Gemini app, close your laptop or lock your phone. Here's what you can unlock with Spark: Set recurring tasks or triggers: Automatically parse monthly credit card statements to flag new or hidden subscription fees. Teach it new skills: Direct it to check your inbox for ongoing updates from your kids' school, extract critical deadlines and send a consolidated daily digest to you and your partner. Create complete workflows: Ask it to synthesize raw meeting notes across emails and chats, create polished Google Docs with its findings and even draft the companion email kicking off a project. Google added that it will continue to give Spark new bells and whistles this summer, including the ability for it to spend your money. This is similar to agentic Gemini, which can already control apps for you. For example, you will be able to ask AI tool to book a ride for you, and it will complete all relevant steps. However, "it's designed to ask you first before performing high-stakes actions like spending money or sending emails." Here's what Spark's rollout could look like: Rolling out to "trusted testers" this week. Rolling out as a beta for US-based Google AI Ultra subscribers next week. Rolling out to the Gemini desktop app this summer.

Google's new Spark AI agent will run your digital life for $100/month

Spark can perform complex functions like drafting emails and creating study guides, with a new Android Halo interface launching later this year to monitor its activities. As predicted, Google has unveiled a Gemini-powered personal AI assistant that will work tirelessly in the cloud, and now we have a name for it -- and no, it won't come cheap. Previous chatter had the Gemini personal AI agent going by the internal moniker of "Remy," but during Google's big I/O reveal on Tuesday, we got the assistant's final name: Spark. Rolling out "deliberately" to "trusted testers" this week, with a wider release to Google AI Ultra subscribers next week, Spark is a 24/7 cloud-based AI agent that "navigates across your digital life." During a pre-brief session, Google reps described how Spark will connect to all your core Google services and documents, including Gmail, Docs, Sheets, and Slides. Give Spark a complex task and it will set off on its own to complete it, spawning sub-agents to tackle different aspects of its assignment as needed, according to Google. (Google hasn't yet fully detailed how Spark's permissions and approval controls will work.) The examples of tasks that Spark could perform include checking your inbox for messages from clients or customers, building student study guides that update themselves automatically as new assignments come in, and writing email drafts based on data from Gmail, Google Docs, and other documents in your Google account. Aside from its access to Google services, Spark will be getting a "whole set" of MCP (Model Context Protocol) connectors for third-party service providers, including Adobe, Asana, Box, Canva, Dropbox, HubSpot, Intuit, Monday, Pandora, Spotify, and Wix. That means Spark will eventually be able to access your Dropbox, check the status of Monday projects, or lend a hand with your Canva creations. Spark will live in the cloud rather than on your desktop PC, similar to Claude Cowork. That means Spark won't be able to directly access files on your system, but it also means you won't have to worry about Spark poking into personal or sensitive documents (like bank statements) in your various local directories. As with other personal AI agents we've seen, you'll be able to communicate with Spark via text and email, meaning you won't need to fire up an app to check in with the assistant. You could also include Spark in text chains or CC it on email threads. Along with Spark comes a revamp of Google's pricier AI Ultra plan, which will now start at $100 per month with a "top tier" capping out at $200 per month (down from the previous single-tier $250 AI Ultra rate). The bottom line, though, is that you'll need to cough up at least $100 monthly to have Spark at your side. (Sorry, AI Pro users.) After Spark's initial launch, Google plans to roll out a new UI space called Android Halo that'll let you view live updates on Spark's progress across its various tasks. Look for Android Halo to arrive "later this year."

Google unveils Gemini Spark -- a '24/7 personal AI agent'

Google has unveiled a new personal AI agent called Gemini Spark at Google I/O today, and it could be the tool that pushes agentic AI into the mainstream. The new tool runs on dedicated virtual machines on Google cloud and is available 24/7 to help "navigate your digital life." The point of agentic AI is to have an assistant that doesn't just respond to your prompts, but takes action on your behalf. But while opening up your desktop to OpenClaw isn't something everyone is willing to (or should) do, Gemini Spark is different. For starters, Gemini Spark will be integrated with Google's suite of products like Gmail and Google Docs that billions of people use around the world every day. The second, and more crucial aspect, is that Google has built Gemini Spark using the Google Antigravity development platform. This is an AI-native IDE (Integrated Development Environment) that prevents the agent from going rogue. So users can effectively send it off to work in the background on tasks that involve agentic coding, knowing that it can be configured to adhere to unbreakable laws. While Spark will launch initially with Google's first-party tools, in the coming weeks, it will also be able to integrate with third-party tools via MCP (Model Context Protocol). If you're not familiar with MCP, it's the open standard that will allow Spark to operate on your behalf with non-Google services, like GitHub or Notion. What can Gemini Spark do? Because Gemini Spark runs in the cloud, it can be accessed at any point from whichever device you're using. You could start interacting with it across Gmail from one of the best Android phones and pick it up later in Google Docs when you get to your laptop. And because it's on at all times, you can set it a task to complete while you're away from your screen -- taking a call or getting lunch. For example, if you use Google's products in your workplace, you could prompt Gemini Spark to pull information from Google Sheets, Docs or Drive and use it to compose an email to a colleague in Gmail, giving them a status update. Or, if you're a student using Gemini Spark on campus, it could work to update your personal study guide in real-time as new assignments arrive from your professor. Here's a few more actions that, Google says, Gemini Spark will be able to do on your behalf: * Set recurring tasks or triggers: Automatically parse monthly credit card statements to flag new or hidden subscription fees. * Teach it new skills: Direct it to check your inbox for ongoing updates from your kids' school * Create complete workflows: Ask it to synthesize raw meeting notes across emails and chats, create polished Google Docs with its findings and even draft the companion email kicking off a project. Google says it will be connecting Spark to a ton of third-party services in the future, including services like Doordash, Asana, Dropbox, CapCut and Uber. While Spark won't launch on Chrome, Google says it will add it to the browser in time. Meanwhile, Android users will also be able to view live updates on tasks assigned to Spark through a dedicated UI space called Android Halo. When will Gemini Spark be available? Google says it is taking a phased rollout to Gemini Spark, starting with trusted testers this week. That will follow as a beta for U.S.-based Google Ultra subscribers next week. Unfortunately, we don't yet know when Spark will come to other Google tiers, including AI Pro and AI Plus, as well as free users. Bottom line Agentic AI has been on the roadmap for several years now and all of the big tech companies are chasing a viable product. We've already seen it on the periphery of mainstream usage with OpenAI's Operator and Opera's Agentic AI browser. But this feels like it could be a pivotal moment. Because Google's products are used so widely across the web for both working and creating, the company is expertly poised to introduce an AI agent into tools millions are already familiar with. I'd bet your garden-variety web user probably doesn't want to install OpenClaw, hand over the keys to their calendar and walk away. But that same user will probably be happy to have Spark reach out to their contacts, agree on a date to connect, find a venue via Google Maps, and make a note in the Google Calendar for them. Of course, Spark is just one part of Google's overall AI game plan and you can check out everything else the company has announced today on our Google I/O live blog right here. Follow Tom's Guide on Google News and add us as a preferred source to get our up-to-date news, analysis, and reviews in your feeds. Subscribe to Tom's Guide on YouTube and follow us on TikTok. Finally, you can visit our dedicated Tom's Guide Savings Squad hub for expert help on getting the best products for less.

Google is giving Gemini far more control over your Mac than before

Gemini Spark will work like a full desktop assistant. Instead of simply answering questions, it is designed to handle repetitive tasks across your Mac, work with local files, and automate workflows that would normally require jumping between multiple apps manually. It can use information from connected apps, conversations, scheduled tasks, browsing activity, location data, and other personalized context to better understand what users are working on. This would allow the AI agent to do things like sort through emails, pull details from documents, complete routine online tasks, or manage workflows spread across multiple apps and services. So, Google wants Gemini Spark to handle more of the repetitive digital work happening behind the scenes, so you can spend less time clicking through tabs and menus.

Google's new AI agent can draft your emails, monitor your inbox and eventually spend your money

Google on Tuesday unveiled Gemini Spark, a personal AI agent designed to work around the clock -- drafting emails, assembling documents, monitoring inboxes, and eventually making purchases -- even when a user's laptop is closed and their phone is locked. The announcement, made at Google I/O 2026, is the company's most ambitious attempt yet to transform its AI assistant from a tool that answers questions into one that autonomously completes tasks. It also arrives at a moment of extraordinary competition, as Microsoft, OpenAI, Anthropic, and Apple all race to build AI systems that don't merely converse but act -- completing multi-step workflows with decreasing human supervision. "We are in that part of the cycle where people want to see real value in the products they use on a day-to-day basis," Sundar Pichai, CEO of Google and Alphabet, said during a press briefing ahead of the keynote address. With Spark, he argued, that value comes from an agent that never stops working. It operates around the clock in Google's cloud, he said, so "you don't need to keep your laptop open to make sure it's running." The product arrives at an inflection point for the technology industry, as Google, Microsoft, OpenAI, Anthropic, and Apple all race to build AI systems that don't merely converse but do -- completing multi-step workflows with decreasing human supervision. It also raises urgent questions about trust, spending guardrails, and what happens when an artificial intelligence agent misinterprets a user's intent. Spark will begin rolling out this week to a small group of trusted testers, with a beta planned for Google AI Ultra subscribers in the United States next week. Inside the cloud architecture that lets Gemini Spark work while you sleep Unlike conventional AI assistants that activate only when prompted, Gemini Spark is architecturally different. It runs persistently on Google Cloud infrastructure, powered by the company's new Gemini 3.5 Flash model and what Google calls the Antigravity agent harness -- the same underlying system that powers the company's internal developer tools. In practical terms, this means Spark can accept a complex instruction -- "email my boss a status update pulling the latest figures from our shared spreadsheet and the project timeline in our Slides deck" -- and then execute it across multiple Google applications without further input. The agent can pull context from emails, documents, and calendar entries, synthesize the information, and produce a finished output. Josh Woodward, VP of Google Labs, Gemini App, and AI Studio, described the experience in visceral terms during the briefing: "When you use it, it almost feels like you're tossing things over your shoulder -- Spark's catching them and gets the job done." The cloud-based architecture is a deliberate design choice. Because Spark operates on remote servers rather than on a user's device, it can continue working through tasks after a user walks away. A student could ask Spark to build a study guide that updates itself as new assignments arrive from a professor. A small business owner could instruct it to monitor their inbox and flag potential customer inquiries. A parent could delegate the logistics of a neighborhood block party -- tracking RSVPs, coordinating contributions, scouting venues. These are not hypothetical scenarios. Woodward said they reflect how early testers have actually been using the product. Over the coming months, Google plans to expand Spark's capabilities significantly. The company will roll out MCP (Model Context Protocol) connections to more than 30 third-party partners, including Canva, OpenTable, and Instacart. Users will also be able to text and email Spark directly, create custom sub-agents for specialized tasks, and connect Spark to Chrome for web-based actions. Later this year, a new Android interface called Android Halo will provide live, at-a-glance visibility into what Spark is working on, displayed at the top of a user's phone screen. Google compares its AI spending safeguards to giving a teenager their first debit card For all its ambition, Spark confronts a fundamental challenge that has bedeviled every AI agent to date: How do you trust an autonomous system to act on your behalf -- particularly when money is involved? Google is acutely aware of the concern. When asked during the press briefing how Spark would avoid making unauthorized purchases, Woodward reached for an analogy that was striking in its candor. "On the team, we think a lot of it is like if you're giving a teenager their first debit card -- there's sort of limits and sort of constraints around it, and that's how we'll be designing Spark as we go through the year," he said. At launch, Spark will not autonomously make purchases. Users will be given explicit opportunities to review and approve any transaction before it goes through. But Google has built the infrastructure for a more autonomous future. Vidhya Srinivasan, who leads Google's ads and commerce teams, introduced the Agent Payments Protocol, or AP2 -- a system designed to let AI agents make secure purchases within user-defined boundaries. The concept works like this: a user tells their agent the specific brands, products, and spending limits they're comfortable with. If the criteria are met, the agent can automatically complete a purchase. AP2 creates what Google describes as a transparent, verifiable link between the user, the merchant, and payment processors, using privacy-preserving technology and tamper-proof digital mandates to ensure the agent is acting within its authorization. AP2 also generates a permanent digital paper trail, so that if a return is needed, the user and the merchant are looking at the same record. Google plans to bring AP2 to its products in the coming months, starting with Gemini Spark. The system is underpinned by the Universal Commerce Protocol (UCP), an open-source standard Google announced earlier this year that gives agents and commerce systems a common language across the entire shopping journey. The UCP Tech Council now includes Amazon, Meta, Microsoft, Salesforce, and Stripe -- a remarkable coalition that underscores how seriously the industry takes the prospect of agent-driven commerce. Google also announced the Universal Cart, an intelligent shopping cart that works across merchants and Google services. Users can add items while browsing Search, chatting with Gemini, watching YouTube, or reading Gmail. The cart then works in the background -- tracking price drops, surfacing deals based on payment card perks, and even flagging product incompatibilities. The shopping infrastructure is rolling out in the U.S. this summer across Search and the Gemini app, with YouTube and Gmail to follow. How Google, OpenAI, Microsoft, Anthropic, and Apple are racing to build the definitive AI agent The announcement lands in the middle of the most intense competitive period in AI history. Google, Microsoft, OpenAI, Anthropic, and Apple are all racing to ship autonomous agents that can do real work -- and each is placing a fundamentally different architectural bet on how to get there. OpenAI recently unified its Operator and deep research capabilities into ChatGPT agent -- a system that brings together website interaction, information synthesis, and conversational intelligence. It carries out tasks using its own virtual computer, shifting between reasoning and action to handle complex workflows. The company emphasizes that users remain in control, with ChatGPT requesting permission before taking consequential actions. But the product has faced scrutiny over reliability. OpenAI's Computer-Using Agent scores 38.1% on OSWorld, the industry benchmark for computer use tasks, while humans score over 72%. Anthropic launched its Claude Computer Use Agent in research preview in March, giving Claude the ability to see, navigate, and control a user's desktop -- clicking buttons, opening applications, filling spreadsheets, and completing multi-step workflows. Claude Cowork handles tasks autonomously -- users give it a goal and Claude works on their computer, local files, and applications to return a finished deliverable. Anthropic has iterated aggressively, recently shipping ten pre-built financial agents and pursuing deep Microsoft 365 integration. Microsoft introduced Copilot Cowork to move beyond chat and into execution -- helping users delegate real tasks and have them completed. Cowork runs in the cloud, meaning users don't have to worry about closing their laptop. The system is grounded in Work IQ, Microsoft's intelligence layer that understands organizational data, tools, and structure. The shift moves Copilot from a sidebar helper to an orchestrator of autonomous agents. Apple is also preparing a revamped Siri for WWDC 2026 that will act as an "always-on agent" capable of handling tasks across apps using personal data. Google's Gemini models will help power the upgraded Siri through a multi-year deal reportedly costing Apple around $1 billion per year. The convergence is unmistakable: every major platform is moving from assistants that talk to agents that act. But each is approaching the problem differently. OpenAI's agent operates primarily through a browser. Anthropic's works directly on a user's desktop. Microsoft's is tightly bound to the Office 365 ecosystem. Apple's emphasizes on-device processing and privacy. Google's approach with Spark is distinctive in its bet on cloud persistence and deep integration with its own services. Rather than controlling a user's screen pixel by pixel, Spark works through structured integrations -- Google's own Workspace APIs, and increasingly, third-party connections through MCP. The advantage is reliability and speed: structured tool use is far more predictable than screen-reading. The disadvantage is that Spark, at least initially, can only act within the systems it's been connected to. The AI model behind Spark processes trillions of tokens a day -- and Google says it could save enterprises billions Spark's capabilities are inseparable from the model that drives it. Gemini 3.5 Flash, also announced Monday, is Google's new workhorse AI model -- designed specifically for the demands of agentic workflows. The performance claims are important. Google says 3.5 Flash outperforms its previous frontier model, Gemini 3.1 Pro, across nearly all benchmarks, while running four times faster than comparable frontier models in terms of output tokens per second. An even more optimized version, available within Google's Antigravity development platform, runs twelve times faster. Pichai framed the economics bluntly. Companies processing roughly one trillion tokens per day on Google Cloud -- a figure he said top enterprise customers are hitting -- could save over $1 billion annually by shifting 80% of their workloads to a mix of Flash and frontier models like 3.5 Pro. In a market where, as Pichai noted, CIOs are already "blowing through their annual token budgets and it's only May," the cost argument may matter as much as the capability argument. Internally, Google's own developers have been consuming Gemini 3.5 Flash at a staggering and rapidly accelerating pace. In March, Google was processing about half a trillion tokens per day internally. That figure has since grown to more than three trillion -- doubling roughly every few weeks. Pichai described this as a "powerful feedback loop" that continually improves the model. Koray Kavukcuoglu, CTO of Google DeepMind and Chief AI Architect for Google, said the model's speed is what makes agentic use cases practical. "3.5 Flash is especially good when deploying multiple agents simultaneously and completing long-running tasks," he said during the briefing, adding that Google had successfully tested agents building "a working operating system entirely from scratch." The 3.5 Pro model, the more powerful sibling, is currently being tested internally and will roll out next month. What Gemini Spark costs and where it fits in Google's new subscription tiers Gemini Spark will be available to Google AI Ultra subscribers. The company is simultaneously restructuring its subscription tiers to make the technology more accessible. A new Ultra plan at $100 per month provides a 5x higher usage limit than the Pro plan, along with priority access to Antigravity and 20TB of cloud storage. The top-tier Ultra plan drops from $250 to $200 per month, with a 20x higher usage limit and access to the full suite of capabilities. Both tiers include Gemini Spark, the Daily Brief agent -- a proactive morning digest that triages email, calendar, and tasks overnight -- and access to the new Gemini Omni and 3.5 Flash models. The pricing positions Spark as a premium product -- more expensive than Anthropic's Claude Pro at $20 per month, but comparable to the higher tiers of competing products like Claude Max ($100-$200/month) and OpenAI's ChatGPT Pro ($200/month). Why privacy, reliability, and ecosystem lock-in could undermine Google's agent ambitions The risks are real and multidimensional. Reliability remains the industry's greatest challenge. Even the best AI models hallucinate, misinterpret instructions, and make errors that a human would never make. An agent that drafts an email to the wrong person, misreads a spreadsheet figure, or sends a payment to the wrong merchant could create consequences that are difficult to reverse. Google's approach of requiring explicit approval for high-stakes actions like spending money or sending emails is a sensible safeguard -- but it also limits how autonomous the agent can actually be. An agent that asks for confirmation at every turn isn't much of an agent at all. Privacy is another concern. Spark's ability to synthesize information across a user's entire Gmail inbox, calendar, documents, and chat history means it has an extraordinarily deep view of a person's digital life. Google says Spark operates on a fully managed, secure runtime with isolated ephemeral virtual machines, encrypted credentials, and Data Loss Prevention policies. But the concentration of personal context in a single AI system -- accessible through natural language -- creates a surface area that will attract scrutiny from regulators, privacy advocates, and security researchers. Market timing is uncertain, too. The consumer appetite for always-on AI agents is unproven at scale. Google says the Gemini app has 900 million monthly users, but it's unclear how many of those users are ready for the conceptual leap from "ask a question, get an answer" to "delegate a task, trust the outcome." The history of digital assistants -- from Clippy to early Siri to Alexa -- is littered with products that promised proactive intelligence and delivered frustration. And then there is the question of ecosystem lock-in. Spark works best within Google's own services. While MCP connections to third-party apps will broaden its reach, the initial experience is one of deep Workspace integration. For the billions of people who live inside Google's ecosystem, this is a natural fit. For those who split their digital lives across Microsoft, Apple, and other platforms, Spark's utility will be more limited -- at least initially. Woodward acknowledged as much when asked whether Spark would remain confined to the Google ecosystem. "It's going to be cross-platform in two ways," he said -- through MCP integrations with third-party apps, and through availability on the web, Android, and iOS, with tasks syncing across devices via the cloud. The real test for Gemini Spark isn't whether it can do the work -- it's whether people will let it Google's bet with Gemini Spark is that the AI industry's center of gravity is shifting from models that think to systems that act -- and that the company best positioned to win that transition is the one with the most comprehensive set of consumer services to act within. It is a bet backed by enormous infrastructure investment. Google expects to spend approximately $180 to $190 billion in capital expenditure this year -- roughly six times what it spent in 2022 -- much of it on the AI compute required to run agents like Spark at scale for hundreds of millions of users. The technology, in other words, is arriving. The models are fast enough, the integrations deep enough, the payment rails secure enough. Google has built a system that can draft your emails, organize your calendar, monitor your inbox, and soon enough, spend your money -- all while you sleep. But the hardest problem in artificial intelligence has never been making a machine capable. It has been making a human comfortable. For two decades, Google's core promise has been ten blue links and a search box -- a transaction built on the assumption that the user is in control. Gemini Spark asks users to renegotiate that relationship entirely, to hand a set of keys to a system that is brilliant, tireless, and still, by its maker's own admission, best compared to a teenager with a debit card. Gemini Spark rolls out to trusted testers this week, with a broader beta for U.S. Google AI Ultra subscribers expected next week.

Google I/O 2026: Gemini Spark is a wildly ambitious AI agent

Google announced Gemini Spark at I/O 2026 today, and it might be the most ambitious thing the company has put on stage. Billed as "your personal AI agent," Spark is a cloud-based AI that runs continuously. According to Google, Spark runs in the Cloud, so you don't need a Mac Mini or a dedicated home server. Plus, it's designed to handle the kind of tasks that pile up in your inbox and to-do list while you're busy doing everything else. Powered by Gemini 3.5 and built with Google's Antigravity coding IDE baked in, Spark connects to Google's own products and integrates with more than 30 third-party tools via MCP, including Adobe, Asana, Dropbox, Lyft, OpenTable, Uber, Zillow, and Zocdoc. Need to send your boss a status update? Spark digs through your inbox and Docs to pull together everything relevant. Running a small business? You can set it to monitor incoming customer inquiries and flag the urgent ones. Google said Spark will land inside Gmail and Chat soon and begin rolling out to AI Ultra subscribers in the U.S. next week. The question everyone has about an autonomous AI agent is obvious: what stops it from doing something you didn't want? Google's answer was the Agent Payments Protocol (AP2), a framework that imposes hard limits on what Spark can spend, which merchants it can interact with, and what it can actually purchase. For now, users approve any transactions before they go through. Josh Woodward, VP of Google Labs, described the philosophy at a media briefing ahead of the event: "On the team, we think a lot of it as if you're giving a teenager their first debit card. There are sort of limits and constraints around it, and that's how we'll be designing Spark as we go through the year." As Google sees it, Spark will gradually gain autonomy over time. For now, though, AP2 will include a permanent digital paper trail for returns and disputes and head to Google Shopping later this year.

Gemini Spark screenshots reveal an AI that actually gets things done on its own

I've been covering Android and other mobile technology for close to ten years now, with a specific interest in phone accessories, e-readers, and what makes each individual phone different from another. I delight in looking at the phone market from as many angles as possible, and while my opinions may be odd, at times, they're always from the heart as much as the head. I have a background in the mobile accessories world, which explains my odd enthusiasm for cases and things that clip onto smartphones. I worked for Digital Trends from 2017 to 2025. It looks like Google is ready to take Gemini to the next level, as early implementations of Google's new AI agent have been leaked. Essentially a semi-autonomous agent in the vein of Claude's Cowork, Gemini Spark looks like it's in close to a finished state, and is likely to be shown off at Google I/O next week. Gemini is coming to eat Claude's lunch We got our first look at Gemini Spark yesterday, as reported by Android Authority. However, then it was nothing but a name on an onboarding screen. Today, however, Spark has been revealed in all its glory as exactly what we'd assumed it was -- an AI agent that can get on with tasks without continual prompting. Screenshots of Spark's capabilities have been posted on X (formerly Twitter) by leakers and testers, and they show some of the agent's pre-built capabilities, as well as its skill creation menus. Close The option to access Spark seems to live in Gemini's hamburger menu opened in the top-left of the app, and when first booted up, it shows a few pre-built ways it can help you out. These include the ability to declutter your email inbox by unsubscribing from unread newsletters or lists, summarizing anything that sounds helpful, as well as monitoring for pre-briefs before meetings, and creating a daily news digest for your interests. You can also create a new skill by supplying it with instructions, while other screenshots show Spark in action. It seems as if Spark will be able to grab information from multiple apps at once, bringing it all together. As you might expect, Google Workspace apps are front and center in a lot of the screenshots, but it's possible it will work with third-party apps given time. It's expected that Spark will include the ability to complete tasks without human input, though unlike Claude's Cowork and OpenClaw, it isn't expected that Spark will have access to the entire computer it's working on. However, it is expected to be able to control browsers, like Chrome. Subscribe to our newsletter for Gemini Spark analysis Want deeper context on features like Gemini Spark? Subscribe to our newsletter for hands-on breakdowns, side-by-side comparisons with other AI agents, and clear analysis that explains capabilities and implications across AI and tech. Get Updates By subscribing, you agree to receive newsletter and marketing emails, and accept our Terms of Use and Privacy Policy. You can unsubscribe anytime. It's very likely that Spark will be a big part of Google's I/O keynote next week, so tune in to see what Google has to say about its new AI agent.

Google's new Gemini Spark AI agent can run your errands while you run your life

Google finally built an AI assistant that keeps working even when you stop. Google just announced Gemini Spark at Google I/O 2026, and it might be the most genuinely useful AI feature the company has ever shown off. Unlike most AI tools that wait on you, Gemini Spark takes a task and runs with it, handling multiple steps in the background without you having to babysit it. The biggest benefit of Gemini Spark is that it runs on dedicated virtual machines, so once you assign a task to it, you don't have to keep your laptop or computer open. You can close it and move away from the desk while Gemini Spark executes the task in the background. What makes Gemini Spark different from every other AI tool? Gemini Spark is powered by the latest Gemini 3.5 model and Google's Antigravity harness. They allow Gemini Spark to run longer tasks in the background. Recommended Videos When you ask Gemini Spark to perform a task, it breaks it down into steps, works across your apps, and handles everything in the background without you having to stay involved. You hand it a job and walk away. It can pull information from your emails, documents, and chats simultaneously, so it always has the full picture. It can draft content, use custom "skills" you upload, update files in real time as new information comes in, and manage follow-ups on your behalf. Currently, Gemini Spark works only with Google's in-house apps (think Gmail, Drive, Docs, etc.), but in the future, Gemini Spark will also integrate with third-party tools, allowing users more freedom. When will it be available? Google is rolling out Gemini Spark to trusted testers first, followed by a beta release for Google AI Ultra subscribers. Google has also introduced a new AI Ultra plan for $100/month to make it more accessible. It's also dropping the price of its premium AI Ultra plan from $250/month to $200/month. Later this year, Spark will also work directly inside Google Chrome as a browser agent, and Google is building a dedicated home for agents on Android called Android Halo. It is still early days, and Google has been careful to say as much. But if Gemini Spark delivers on even half of what the demo showed, it could finally be the AI assistant worth getting excited about.

Google Launches Gemini Spark: A 24/7 AI Agent That Wants to Make You Ditch OpenClaw - Decrypt

Spark integrates with Gmail, Docs, and Slides out of the box, and adds MCP connections to Canva, OpenTable, and Instacart starting today. Google's Gemini app has more than 900 million monthly users. Up until now, most of them used it as a very smart chatbox. That changes today. At Google I/O 2026, the company announced Gemini Spark, a personal AI agent that doesn't wait to be asked. It manages tasks across your apps around the clock, flags things that need your attention, and finishes jobs in the background while you're busy doing literally anything else, including sleeping. Spark runs on Gemini 3.5 and is built on Google's Antigravity harness -- the same agent infrastructure behind the company's own internal tools. Unlike OpenClaw, Spark lives in the cloud, on dedicated Google virtual machines, so you don't need your phone unlocked or your laptop open for it to keep going. Out of the box, Spark can handle recurring tasks and do most of the everyday tasks you have handled to your Openclaw or Hermes agent. It can automatically scan your monthly credit card statements for new or hidden subscriptions. It can monitor your kid's school emails, pull out important deadlines, and send you and your partner a daily digest. It can pull raw notes from a meeting scattered across your Gmail and Docs, write them up into a clean document, and send the follow-up email to kick off the next phase of a project. You can also teach it custom skills -- essentially programming it with your preferences so it handles specific workflows the way you want them handled. None of this was a complete surprise. Last week, developers digging through the Gemini Android app's latest beta build found the "Gemini Spark" name -- previously labeled internally as "Gemini Agent" -- along with a welcome screen that described the same use cases Google announced today. Some internal references also called the project "Remy." The leaked code also included a disclaimer Google is keeping in the official release: Spark "may do things like share your info or make purchases without asking." The company says it's designed to request permission before high-stakes actions, but users are advised to supervise it. At launch, Spark works natively with Gmail, Google Docs, Slides, and the broader Google Workspace suite. Starting today, it also gains MCP connections to Canva, OpenTable, and Instacart -- meaning it can start doing things like booking a restaurant or placing an order, not just drafting the message about it. Decrypt wasn't able to test the agent right now to see if it supports custom MCP configurations, because it's still in limited access. Over the summer, Google plans to add the ability to text and email Spark directly, create custom sub-agents, and let it operate your local browser. A macOS desktop version is also coming, which will extend its reach to local files. The broader AI agent wave has been building for months -- open-source tools exploded and self-improving agents started going viral on GitHub. Spark is Google's answer to all of it: no terminal, no setup, no GitHub stars required. Just 900 million existing users, a deep integration with tools they already rely on, and an agent that gets to work the moment you turn it on. Trusted testers get access this week. U.S. Google AI Ultra subscribers -- whose plan Google is cutting from $250 to $200 a month -- get beta access next week.

Why Google's Gemini Spark AI agent could be a game changer

Megan Cerullo is a New York-based reporter for CBS MoneyWatch covering small business, workplace, health care, consumer spending and personal finance topics. She regularly appears on CBS News 24/7 to discuss her reporting. Google's Gemini Spark AI tool could be a game-changer for consumers wading into the world of agentic AI, according to technology experts. Google announced the new tool -- the internet giant's answer to OpenClaw, the autonomous AI agent being hyped by tech enthusiasts and developers -- this week during a rollout of new AI features that Google said can complete tasks for users. "There is a paradigm shift happening right now where AI is going from a chat interface to actually being able to do things for you," Clarence Lee, a tech entrepreneur and visiting lecturer at the Cornell SC Johnson College of Business, told CBS News. Lee compared an AI agent like Gemini Spark to a personal assistant to whom you can assign tasks, such as composing emails, making restaurant reservations or booking hotels. "It's like having a team that you can delegate things to," he said. For an individual AI agent to offer optimal performance, it must understand its human users, said Karan Girotra, professor of operations, technology and innovation at Cornell University. "You give it an objective, and it takes actions to accomplish that on your behalf," he told CBS News. "For it to be good, it needs intelligence, context and information relevant to those actions." That's where Google could have an edge on competitors. Gemini Spark is integrated into Google's widely used suite of digital tools, including Gmail and Google Calendar, giving the AI access to important contextual information about a user. "It knows more about you than many others because it connects to Gmail and other apps, so personal intelligence will come through in the agent," Girotra said. Google on Tuesday said it is testing Gemini Spark this week and rolling it out to Google AI Ultra subscribers, who pay $100 a month for the highest level of access to its AI tools, next week. "Spark represents a big shift for Gemini, transforming it from an assistant that can answer your questions into an active partner that does real work on your behalf and under your direction," Google said in a blog post. Gemini Spark syncs up with Gmail, Docs, Slides and other Google apps. With access to your inbox, it can comb through your emails and alert you to updates from your kids' school, or remind you about key deadlines, Google said. The agent can also synthesize your meeting notes into a polished document highlighting key takeaways, Google said. Gemini Spark can complete tasks outside Google's ecosystem by connecting to external tools such as Instacart and OpenTable, according to the Alphabet-owned company. That enables users to ask Gemini Spark, for example, to make a dinner reservation or buy groceries. Google said Gemini Spark will need to obtain a user's permission before executing "high-stakes actions like spending money or sending emails." Cornell's Lee also advised moving gradually into the world of AI agents by initially entrusting them with low-level tasks. "The first time you onboard an assistant, you don't know how good they are, so you try them out a little bit before you hand over your credit card," he said. "You might have them draft emails or create a grocery list, so I recommend that users start that way."

Gemini's Spark agent has leaked, and it looks like it's gunning for Claude Cowork's throne

Spark is expected to interact with other apps to execute multi-step tasks without human intervention. Earlier this week, Google showcased the facelift that's coming for Gemini, along with new automation features. But along with this, Gemini may be getting a truly agentic upgrade with an always-active sidekick that lives in your chats and performs tasks for you, rather than just responding to queries. After an early preview yesterday that revealed this agent could be called Gemini Spark, we have more leaks detailing its functionality. We've come across reports from several tinkerers on X, including @Waguri_Kaoruko8 and @testingcatalog, who have enabled Spark within Gemini. According to the reports, the option to activate Spark appears in the overflow menu, accessible via the hamburger (three horizontal bars) button in the top-left corner of the Gemini launcher on Android. The welcome screen elaborates on different ways Spark can help you, and it involves performing actions in other apps. For instance, it can clear up junk from your Gmail inbox, assemble notes before key meetings, or create a custom news digest. However, that's not all it can do. Based on other screenshots, you will be able to create specific skills, which might allow you to complete recurring tasks with a similar set of instructions, but with some variable that can be fed through the prompt. Claude offers a similar functionality through Projects. Besides this, Spark should be able to perform multi-step workflows, indexing information from multiple apps simultaneously. While the screenshots currently only show Google Workspace apps being summoned, we might see support for third-party apps as well. Based on what we've seen earlier, users will have the option to let Spark function without a human reviewing its work. Additionally, Spark might even rely on a separate AI model for its functioning. Testing Catalog also notes that Spark should be able to control the Chrome browser like an agent and use files stored on your computer or other devices, but it may not yet be able to control the entire computer, like OpenClaw and Claude Cowork. While Google hasn't officially confirmed anything about the agent, an announcement is highly likely at the Google I/O 2026, scheduled for next week.

Google unveils Gemini Spark, an always-on AI agent for daily digital tasks - SiliconANGLE

Google unveils Gemini Spark, an always-on AI agent for daily digital tasks Entering into the AI agent era, Google LLC today introduced Gemini Spark, a 24/7 personal artificial intelligence assistant that can help people navigate their digital lives and do real work on their behalf. Under the hood, Spark runs on the newly released Gemini 3.5 Flash and uses the company's updated Antigravity platform to orchestrate AI agents. It is deeply integrated into the company's ecosystem of products that users interact with daily, including Gmail, Docs, Slides and more -- allowing the agents to find information, understand context, learn personalized daily routines and help users catch up. Most importantly, Spark continues working in the background even while the user wanders away from their laptop or phone. They can set it up like a digital secretary to take complex goals off their plate and handle tasks such as cleaning up their inbox, searching for information and keeping a lookout for things that might otherwise get lost in the noise. With Spark, users can set triggers, such as monitoring credit card statements to warn about new subscription fees or hidden costs. Users could also direct the agent to watch their inbox for ongoing updates from their work or daily life, helping them stay up to date on colleagues, critical deadlines or creating a daily digest so they can remain on task. Users can even have the agent compile raw meeting notes across emails and chats, allowing them to "stay in the know" by generating Docs files with its findings and drafting companion emails when starting projects. Google said this is only the beginning for Spark. Right now, the assistant connects to the company's ecosystem, but it will soon ship with third-party connectors. Model Context Protocol connections to Canva, OpenTable and Instacart launch today, and Google said more third-party partners will be announced in the coming weeks. New abilities for Spark are also planned, including texting and emailing Spark, creating custom sub-agents and operating the local browser. Spark is a powerful agentic assistant, so Google said users will remain in full control. They pick when to turn it on and what apps it can connect with, and it is designed to ask before performing high-stakes actions such as spending money or sending emails. Google redesigned its Gemini experience and Gemini app to be more intuitive with a new design language the company is calling "Neural Expressive," featuring fluid animations, vivid colors and touch feedback. According to the company, the new visuals are designed to be flowing and visually pleasing so users can see that the interface is "thinking" and engage with it. The app and interface now directly integrate Gemini Live, the voice-based conversational capability. This will allow users to easily switch between text and voice, quickly going from typing long-form questions into Gemini's interface to asking a quick question in a back-and-forth voice conversation -- and then going back again. Google said this change will allow users to approach ideas and conversations with Gemini at their own pace, depending on how they need to express their needs, without being cut off mid-thought. To foster better responses, the company added that it is using the newest model architectures to better understand conversation. For example, instead of throwing an entire wall of text at users each time, Gemini will judge whether to be long-winded, offer a short reply or incorporate rich imagery, interactive timelines, narrated videos and dynamic graphics. Users can also opt into a Daily Brief. Once they do, an agent will brief them with a personalized morning digest designed to be the first stop based on their inbox, calendar, search and other preferences. It will act as a critical update on their day and what someone should know about anything upcoming -- it will go beyond a simple summary, organizing and prioritizing based on specific goals, even suggesting immediate next steps. Users will also be able to steer its behavior with a quick thumbs-up or thumbs-down. Google said it has big plans in particular for the macOS Gemini desktop app. The company intends to bring Spark to the app, where it will be able to handle complex tasks with local files on the desktop, similar to how Anthropic PBC's Claude Cowork operates, and automate work. The company also said it is innovating the desktop app's voice experience. Users do not need to worry about "ums," "ahs," and "how abouts" when thinking aloud. They can just narrate, and using context on the screen, Gemini will take free-flowing speech and turn it into precise drafts, instantly reformatting even scattered thoughts to capture intent right where the cursor is.

Gemini is about to get wings on your phone with agentic skills

Google I/O is almost here, and now that Google has already wrapped up The Android Show, all eyes are shifting toward the company's AI ambitions -- especially Gemini. While nothing has been officially announced yet, a new leak gives us an early glimpse of what Google could be preparing behind the scenes. Your inbox might soon fear Gemini more than spam A post shared by Waguri_Kaoruko8 on X reportedly showcases something called the "Gemini Spark Model," alongside a new Agent or Chat Mode designed for more advanced tool-based actions. And honestly, this feels like Google trying to turn Gemini into a proper AI assistant that handles annoying digital chores for you. Recommended Videos The leak was later quote-posted by the AI News testing catalog account, which claimed there currently doesn't seem to be support for importing SKILL MD files directly. Apparently, users may have to rely on the good old copy-paste method for now. The post also mentions there's no sign of browser control or full computer-use capabilities yet -- two features many people were hoping would eventually arrive as AI agents become more capable. But the screenshots themselves are where things get interesting. According to the leaked interface, the Gemini Spark Model -- currently labeled as beta -- appears focused on automation and personalization. One feature can reportedly clean up your inbox by summarizing newsletters, archiving clutter, and even automatically unsubscribing from mailing lists. Another tool can generate meeting briefs, pulling together relevant information and quick summaries before an important call or appointment. There's also a custom news digest feature that seems designed to follow the stories you actually care about, rather than flooding you with random headlines all day. In a way, it feels like Google is pushing Gemini toward becoming a background productivity layer rather than just an AI you occasionally ask questions to. And frankly, that's probably the smarter direction. Google may be building a DIY AI workflow system The leak also suggests users may be able to create custom "skills" for Gemini. The setup process reportedly involves giving the skill a title, explaining what it does, and adding instructions for how Gemini should behave. Think of it almost like building mini AI workflows without coding. Of course, it's important to keep expectations in check here. None of this is official yet, and leaks around Google I/O season tend to fly around fast. Still, the timing makes this particularly interesting. With Google expected to go all-in on Gemini at I/O next week, there's a very real chance we could see at least some of these features become official sooner rather than later.

Google launches Gemini Spark as a personal AI agent

Google has introduced a new AI assistant named Gemini Spark, described as a "24/7 personal AI agent" that functions as an "active partner" instead of a conventional assistant. The new assistant is powered by Gemini 3.5 and is integrated with Google Workspace applications, including Gmail, Docs, and Slides. Users can teach Gemini Spark to perform various tasks, such as creating lists of deadlines in Gmail and summarizing lengthy email threads. The assistant can also be programmed for recurring tasks, such as identifying hidden fees in credit card bills each month. Additionally, Spark can execute interconnected tasks; for example, it can generate reports in Google Docs based on meeting notes and draft related emails. Video: Google Gemini Spark is rolling out to testers this week and will be available to Google AI Ultra beta users in the United States next week. The usage of Spark is opt-in, allowing users to decide whether to enable the assistant and select which applications it can connect to. In addition to Google Workspace apps, Spark currently supports integrations with Canva, OpenTable, and Instacart, with more partner applications expected in the coming weeks. Google aims to release additional features for Spark in the future, including the capability to send texts and emails and to control the browser. The company stated that Spark will prompt users before performing significant actions, such as spending money or sending emails. Furthermore, Gemini Spark will be incorporated into the Gemini desktop app this summer, enabling it to access files and perform tasks directly on users' computers.

Google's Gemini Spark is the first AI agent built for people who don't know what an AI agent is

Quentyn is a career tech journalist with nearly two decades of experience. He has written thousands of stories for publications such as The Verge, Forbes, Consumer Reports, and Business Insider. Quentyn specializes in reporting about mobile technology (including smartphones, tablets, wearables, and accessories), home theater, gaming, computing, and cameras. In his free time, you can find him abusing any variety of gaming controllers or learning about his latest muses (most recently, drones and cooking). Summary Gemini Spark is a 24/7 cloud-based personal agentic AI that runs complex, long tasks for you. Native Google Workspace and MCP integrations let Spark automate workflows in external apps. You're in control: Spark asks before completing sensitive tasks like making payments. Compared to the explosion of conversational chatbots in recent years, the road to agentic AI has been more like a soft trot than a march. Claude and OpenClaw gave us an early taste of the power this proactive and autonomous evolution of AI can wield, but until now, they've often felt more like a developer's experiment. Google has thrown down the gauntlet as the first major player to change that by announcing Gemini Spark at Google I/O 2026. It's a 24/7 personal AI agent that prioritizes out-of-the-box user-friendliness, seamless app integrations, and entirely cloud-based processing. Gemini Spark is rolling out to trusted testers this week, with a beta launch planned for US Google AI Ultra subscribers next week. Related Google's new "Neural Expressive" AI mode fixes the biggest issue I have with Gemini Google has addressed a massive UX flaw in modern AI apps. Posts By Sara Heritage How Gemini Spark stands out from OpenClaw and Claude It's not just for coders and geeks The biggest barrier to entry for current agentic AIs is the required setup and 0upkeep. If you want an open-source agent like OpenClaw to continually monitor your workflows, you traditionally have to run it on your own hardware or pay to keep a server instance running. Gemini Spark eliminates this hurdle by being entirely cloud-based. Powered by the new Gemini 3.5 model and Google's "Antigravity harness," Spark keeps working in the background even when you close your laptop or lock your phone. You don't have to babysit the AI or monitor server loads; it simply takes complex, long-running, and repeatable tasks off your plate and delivers the results when you return. Getting other agentic models to interact with third-party software often requires coding your own connections and other technical know-how that intimidates the average user. Spark, on the other hand, is built to be intuitive for everyone and works with the same conversational prompting you'd use in standard Gemini. Related Microsoft just announced its answer to OpenClaw, and it actually looks pretty great OpenClaw may be more powerful, but that power comes with major security tradeoffs Posts By Dave Schafer Spark connects with other apps to run your life But don't worry, you'll have full control Spark naturally features deep integration with the Google Workspace tools you rely on daily, such as Drive, Gmail, Docs, and Slides. Google envisions Spark as an active partner that operates under your direction to handle digital chores using tools and resources from various internal and external apps. Some of the practical capabilities Google showcased include: Managing recurring tasks: You can instruct Spark to automatically parse your monthly credit card statements in the background to flag any new or hidden subscription fees. Creating automated workflows: Ask Spark to synthesize raw meeting notes from your emails and chats, generate a polished Google Doc with its findings, and even draft a companion email to kick off the project. Information extraction: Teach the agent to constantly check your inbox for updates from your kids' school, extract the critical deadlines, and send you a consolidated daily digest. Google is also using the new Model Context Protocol (MCP) to enable Spark to connect directly with partner apps and act on your behalf. Initial integrations launching today include Canva, OpenTable, and Instacart, with a massive roster of partners currently planning more. It can actively use these apps to get things done -- like booking dinner reservations or ordering groceries. If that sounds like too much privilege, you'll be relieved to know that you can choose exactly which apps it can connect to, and that it's explicitly designed to ask your permission before executing any sensitive actions, such as spending money or sending emails. Google Gemini OS Android, iOS, macOS, Windows Developer Google Price model Free, Subscription See at Google See at Google Play Store See at App Store Expand Collapse

Google Shows Off Gemini Spark, its New Always-on Personal Agent - Phandroid

Google I/O is finally here, although unlike previous conferences where the company focused on Android and its Pixel devices, this year's show pretty much focused on AI. While Google showed off a lot of new additions to its AI ecosystem, it also unveiled Gemini Spark, which essentially works as a 24/7 personal AI agent. For those unfamiliar, AI agents are a smarter step above your basic chatbot, and in addition to search queries they are also capable of handling tasks independently in the background, such as planning schedules, managing your groceries, ordering take-away, and more. That being said, Spark is powered by Gemini 3.5 and features deep integration with Workspace tools such as Gmail, Docs, and Slides, for example. It's worth mentioning that since Spark operates entirely as a cloud-based agent, it continues executing tasks in the background even after a user closes their laptop or locks their phone. Gemini Spark allows users to set recurring tasks or triggers, such as instructing the agent to parse monthly credit card statements to flag new or hidden subscription fees, for example. The agent can also be taught new skills, such as monitoring an inbox for ongoing updates from a specific sender, extracting critical deadlines, and sending a consolidated daily digest to multiple recipients. Google adds that it's designed Spark to execute complex, multi-step workflows -- for example, it can synthesize raw meeting notes across scattered emails and chats, compile those findings into a polished Google Doc, and even draft a related email.

Google's upcoming 'Gemini Spark' could soon book your flights and handle your inbox

Google has experimented with AI agents before with internal projects like "Remy," but the current Gemini Agent tools have been largely confined to paying AI Ultra users. Spark looks like the next step in that evolution, but this time with a wider consumer-facing angle. Meanwhile, Google is also warning users about the risks. The onboarding screen says Gemini Spark can share sensitive information with third parties or make purchases without asking for confirmation in some situations. The system can also store remote browser session data, including login information and remote execution states, so that workflows can keep running in the background. The Gemini settings will apparently allow users to clear that data or turn off connected services.

Google IO 2026: Google Brings Gemini Spark to Apple's Mac

Gemini Spark was introduced by the Mountain View-based tech giant on Tuesday during its annual developer conference, Google I/O 2026. The new Gemini Spark is the company's latest personal AI assistant, which is backed by Google's latest Gemini 3.5 series models. During the event, the tech giant also announced that its new personal AI assistant will also be rolled out to Apple's macOS in the Gemini app, along with a new voice experience, similar to Google's Rambler. The Gemini Spark upgrade will initially be available in the US to paid subscribers, while other features will be released globally. Google IO 2026: Gemini App on macOS Gets New AI Features Google's latest developer conference witnessed a number of announcements from the tech giant, mostly around the AI upgrades its products are receiving. However, the company also announced that its Gemini app for macOS is being upgraded, too. The dedicated Gemini app for macOS will soon receive the Gemini Spark update, which will allow Mac users to organise local files or extract PDF data directly into Google Sheets and the Gmail app on Apple's desktops and laptops running macOS. On top of this, the company has also announced that the Gemini app for macOS will be upgraded with new "voice experiences", where Gemini will be able to refine speech-to-text inputs. The app will be capable of correcting words and removing unnecessary phrases like "um," "ah", and "like" as the user narrates the text input. The feature seems identical to Rambler, which was introduced during The Android Show I/O Edition last week. However, the tech giant did not confirm whether it is the same AI-powered speech-to-text tool or not. Apart from this, the Gemini app for macOS will soon be able to access the Mac users' screens to retrieve context. Google highlighted that this will let Gemini turn a Mac user's "free-flowing speech into precise drafts", while also reformatting the text to capture the user's intent. Google highlighted that while the Gemini app for macOS is currently available for download for all users, the Gemini Spark upgrade will be rolled out to Mac users with the Google AI Ultra subscription this summer in the US. Meanwhile, the new voice-based AI tools for the dedicated Gemini app will be released globally to all users in the coming weeks.

Google I/O 2026: What is Gemini Spark, the AI assistant that works even when your laptop is off

The company also introduced AP2 safeguards to control spending permissions and AI-driven transactions. During the Google I/O 2026, the company introduced Gemini Spark, a new AI-backed assistant that is made for operating like a personal autonomous agent rather than a traditional chatbot. The company stated that Spark is a cloud-based AI system that is capable of continuously working in the background to help users manage emails, schedules, documents, and everyday workflows. The Gemini Spark is backed by Gemini 3.5 and is integrated with Google services such as Gmail and Chat, while also supporting more than 30 third-party platforms through MCP integrations. It includes apps and services such as Adobe, Asana, Dropbox, Lyft, OpenTable, Uber, Zillow, and Zocdoc. The company says the AI can automatically collect information across connected apps, prepare summaries, track customer requests, and organise tasks without requiring constant manual input. The company has also announced that Spark will use Google's Antigravity coding IDE infrastructure to support more advanced task execution and workflow management. Google also says that the system is designed to reduce repetitive digital work by handling routine organisational tasks in the background. To keep the concerns around AI autonomy and spending permissions, Google has announced a new safeguard framework called Agent Payments Protocol (AP2). The system places limits on the purchases, merchants and transaction limits while needing users to approve payments before completion. Google also stated that AP2 will maintain a permanent digital transaction trail for dispute handling and returns. Google also hinted that Spark's autonomy will gradually expand over time as the company works on safety controls and user permissions. Coming to availability, Gemini Spark is expected to begin rolling out to Google AI Ultra subscribers in the United States next week, with Gmail and Google Chat integration arriving first.

Google reportedly working on Gemini Spark, an always-on AI assistant: Here is what it may do

Gemini Spark is said to be an always-on AI assistant that can complete tasks for users in the background instead of waiting for prompts. Google is currently preparing for its I/O 2026 event, which is scheduled to kick off next week. Ahead of the event, a new leak has revealed details about a possible new AI agent called Gemini Spark. The upcoming tool is said to be an always-on AI assistant that can complete tasks for users in the background instead of waiting for prompts. According to a screenshot shared by X user Fandu and reported by Android Authority, unlike traditional AI chatbots, Gemini Spark may continuously run in the background and handle digital tasks automatically. As per the report, Gemini Spark could be one of the major announcements at Google I/O 2026. Also read: OpenAI Codex now available in ChatGPT mobile app: Features, availability and more The onboarding screen reportedly suggests that Gemini Spark can access data from linked apps, chat history, scheduled tasks, websites where users are already logged in, location information, and Personal Intelligence. With this information, the AI assistant could potentially manage inboxes, complete online workflows, and perform multi-step tasks across different apps and services. Google has previously experimented with AI agents through internal projects like Remy, but many of its advanced AI features have largely been limited to paid AI Ultra subscribers. Gemini Spark could be the company's next major step towards bringing AI agents to more users. The company is also reportedly highlighting the possible risks of Gemini Spark. The onboarding screen warns that Gemini Spark can share sensitive information with third parties or even make purchases without asking for confirmation. Also read: OpenAI may sue Apple over how ChatGPT was integrated into iPhones: Here is what happened The system may also store remote browser session data, including login details and remote code execution data. Google is not alone in building AI agents that can perform tasks on behalf of users. Companies like OpenAI and Anthropic are also developing AI systems capable of browsing the web, managing workflows and completing tasks with minimal user input.