Google turns Gemini into a proactive AI agent with Spark, Daily Brief, and major redesign

Google just announced one of the biggest overhauls to the Gemini app yet at Google I/O 2026, introducing a redesigned interface, proactive AI agents, and a powerful new video generation model called Gemini Omni.

The company claims Gemini is now used by more than 900 million people across 230 countries and territories, making it “the most widely available generative AI tool in the world.” As part of its latest push to make Gemini even more powerful, Google is now evolving the chatbot into a more proactive assistant that can actively help manage your digital life.

The biggest visible change to the Gemini app is a new design language Google calls “Neural Expressive.” The refreshed interface is already rolling out to users and introduces fluid animations, updated typography, vibrant colors, haptic feedback, and a more dynamic presentation style.

Google says Gemini will now generate responses with richer formatting, including images, summaries, bolded text, interactive graphics, timelines, and even narrated videos to avoid showing users giant walls of text.

Google says it is also fully integrating Gemini Live into the core Gemini experience. Users can now switch seamlessly between typing and natural voice conversations without breaking context. The company says it also redesigned the microphone experience so users can speak naturally without interruption, and that support for regional dialects is coming soon.

The Gemini app’s Neural Expressive redesign starts rolling out globally today across Android, iOS, and the web.

At I/O 2026, Google also introduced Gemini Omni, a new multimodal model focused on video creation. Omni has been doing the rounds on the internet lately, and can combine text, images, and video prompts to generate high-quality cinematic videos. Google says users will be able to access Omni straight from the Gemini app.

Users can upload footage from their camera roll and edit it using natural language prompts rather than traditional editing tools. Google also says users will be able to create AI avatars that look and sound like them.

Gemini Omni is also rolling out today to Google AI Plus, Pro, and Ultra subscribers worldwide.

Beyond the Neural Expressive redesign and Omni, Google’s big upgrades to Gemini also include two brand new AI agents — Daily Brief and Gemini Spark.

As the name suggests, Daily Brief is a personalized digest that summarizes your day using information from your calendar, reminders, travel plans, and more. It’s a feature Google previously attempted to launch as Daily Hub on the Pixel 10 series, but removed due to poor user feedback.

Google describes Gemini’s Daily Brief as a morning briefing experience that prepares you for the day ahead. It is launching today inside the Gemini app for Google AI Plus, Pro, and Ultra subscribers in the US.

The second, and more ambitious, feature is Gemini Spark. Google describes Spark as a 24/7 cloud-based AI agent that can proactively handle tasks on your behalf even after you close your laptop or lock your phone.

Powered by Gemini 3.5 Flash, Spark deeply integrates with Workspace apps like Gmail, Docs, and Slides. Google shared examples like parsing credit card statements to identify hidden subscriptions, monitoring school emails for deadlines, creating daily digests for families, and even automatically turning scattered meeting notes into polished Docs and draft emails.

Spark can also create recurring workflows and triggers. Google says users remain in control, and Spark will ask for permission before taking high-stakes actions like spending money. Speaking of which, Google has created a new Agent Payments Protocol (AP2) to help agents like Spark make secure payments on your behalf. The company says it will start bringing AP2 to Google products in the coming months, starting with Spark.

Gemini Spark rolls out to trusted testers this week, with a beta launch for US Google AI Ultra subscribers expected next week.

Google is also expanding Gemini’s connected app ecosystem through new integrations launching today with Canva, OpenTable, and Instacart. Spark will eventually use those integrations to complete tasks across these services. Additional capabilities planned for later this year include Chrome integration, texting and emailing Spark directly, custom sub-agents, and local browser control.

Finally, Google confirmed that its macOS Gemini app is getting deeper Spark integration and new voice capabilities, allowing Gemini to work more directly with local machine workflows.