Google used its annual developer conference on Tuesday to outline an aggressive new vision for artificial intelligence, unveiling a wave of products and upgrades designed to make Gemini the centre of its ecosystem.
At the company’s I/O 2026 event in California, CEO Sundar Pichai described the company’s next era as “agentic,” as the company shifts towards an AI system capable of completing autonomous tasks.
To Pichai, the agentic era also includes AI agents, which will be integrated into many of the company’s products, including Google Search and Gemini, Google’s artificial intelligence (AI) chatbot.
“Ten years since we pivoted the company to be AI-first, we still see AI as the most profound way to advance our mission and improve people’s lives at scale,” Pichai said during his keynote speech.
The announcements come as Google faces mounting pressure from rivals, including OpenAI, Microsoft and Anthropic, in the rapidly expanding AI market.
Biggest search overhaul in 25 years
One of the largest changes for Google is a major redesign of Google Search, the flagship search engine that the company said is used by over 3 billion people.
The Intelligent Search box, powered by AI, will let users explain exactly what they need without having to use keywords in what the company is calling “the biggest upgrade to our Search box in over 25 years.”
A demonstration of the feature on a Google blog shows a user typing directly into the search box that they are looking to start a new hobby in pottery and asking for recommendations for classes on Tuesday nights or weekends near their location.
The user then gets a more targeted list of search results and can ask follow-up questions through Google’s AI mode so “context stays with you as you explore more deeply,” the blog said.
For longer tasks, such as planning a wedding or managing a move, Google is introducing coding agents into search that it said will help users create a custom dashboard or “mini apps” that users can use to come back to and make progress.
Google Search will also have information agents, which are “personalised AI agents” that users can set up to work in the background, “to find what you need at exactly the right moment,” Pichar said.
For example, a user asks Google search to keep them updated for when their favourite athlete announces a collaboration with a shoe brand, and Google will send them a notification for relevant stories that pop up on this anywhere on online sites, blogs, social media posts and news sites.
Google’s search overhaul comes after concerns that AI Overviews, the AI-generated answers in the search bar, are reducing traffic to external websites.
Two new Gemini models
Google used the conference to announce its latest AI tools: Gemini Omni, which can “create anything from any input” and Gemini 3.5 a lightweight version of its AI model that can operate faster and at a lower cost.
Omni will let users combine video, images, audio and text to create high-quality videos, according to the company. It also lets users edit videos through having a conversation with the AI about what to change in the video.
One demonstration for the AI shows a male user taking a video of himself touching a mirror, then his hand becoming metallic, after a prompt asked Gemini to “make the mirror ripple like liquid,” and turn the person’s arm into that material.
The new model is also able to reason about what should happen next in a scene by using a mix of Gemini’s “knowledge of history, science and cultural context,” a blog about the new model said.
Google said Gemini 3.5 Flash, on the other hand, is optimised for developer and auditing work, which can be done in “a fraction of the time” and at “less than half the cost of other frontier models,” because it can solve real-world problems such as developing new applications.
A more advanced Gemini 3.5 Pro model was also teased during the event, although Google confirmed it will launch later this year.
Conversational AI throughout Google products
Google also announced expanded AI integrations across its consumer apps and productivity services.
New features include voice-powered Gmail search, Docs Live, which will let users create documents from their verbal notes, “Ask” YouTube, a search tool that will let users ask questions directly on the video platform, and AI-assisted image editing capabilities.
On YouTube, users will see “videos that best match your interest, and … jumps right to the part of the video most relevant to you.”
The company also introduced updates to its developer platform, including new AI Studio tools that allow users to generate Android applications from natural language prompts.
Google’s push into AI eyewear
Google also renewed its push into extended reality hardware with updated Android XR smart glasses.
The company is developing two types of smart eyewear: audio glasses that offer spoken help in the user’s ear and display glasses that show the user the information live when they need it, the company said.
The first is audio glasses, where the user can either say “Hey Google” or tap the side of the frame to trigger live assistance from the company for more information about what they’re seeing, help with directions, take and edit photos or conduct live translation.
Google confirmed partnerships with eyewear brands including Warby Parker and Gentle Monster, and said the audio glasses will be launched as part of their fall collections.
