Google is reinventing Search with the introduction of Agentic AI

Big Tech company Google has announced the dawn of agentic intelligence at its Google I/O 2026. The company is entering an era where it’s no longer an ‘answering machine’ that only provides answers to prompts, but is becoming a proactive, 24/7, background ‘agentic AI’ system that operates in the background, carrying out actions on behalf of users.

The company maintained that the goal of Search has always been simple: to help users ask anything on their mind — from quick facts to the deep, complex or hyper-specific questions that can be hard to articulate.

“We’re bringing our advanced model capabilities to Search with new AI features, enabling you to use agents just by asking a question,” Liz Reid, VP, Head of Search, said. “We’re also introducing a new, intelligent AI-powered Search box, marking its biggest upgrade in over 25 years.”

Google has continued to reimagine what Search can do with AI, and within one year of introducing AI Mode in search, the company reports surpassing one billion monthly users, with queries more than doubling every quarter since launch.

“As people have realized just how much more Search can do for them, they’re searching more than ever before — so much so that last quarter, we saw queries reach an all-time high.”

Google’s recent AI upgrade is powered by the frontier models: Gemini 3.5 Flash and Gemini Omni Flash.

Gemini 3.5 Flash: The new default engine for Google’s AI ecosystem globally; it enables Google to move beyond generating text into actively completing complex, long-horizon tasks for consumers and developers at a fraction of the cost.

It runs four times faster than competing frontier models in output tokens per second, ensuring users no longer have to trade quality for speed. It outperforms previous models on advanced coding and agentic benchmarks.

Gemini Omni Flash: this is a native multimodal model that democratises cinematic video production. It can generate and edit high-quality video using any combination of inputs: text, images, audio, or video. It enables users to edit video via natural, conversational instructions.

Here’s a breakdown of some of the key updates from the Tech Giants.

With Gemini 3.5 Flash, users get a completely reimagined AI Search box that accepts inputs such as images, text, files, videos, and even Chrome tabs. What this means is that Google is officially repositioning Search from just a reactive link-retrieval engine to a proactive, highly customised platform that builds the specific tools users need.

Search can now code custom “mini-apps” and build interactive visual layouts entirely on the go. An example of what you can build with this is a bespoke expense tracker tailored to your financial goals.

The new AI Search box rollout has started and is available in all countries and languages where AI Mode is available.

Gemini is also transitioning into a 24/7 personal AI agent, making it the clearest example of the “Agentic Era.” With this, users can leverage AI to step in and actively manage digital chores and execute workflows on their behalf, but only under their explicit direction.

Gemini Spark, a cloud-based agent, runs in the background (even with a closed laptop) to execute multi-step workflows, such as parsing credit card statements for hidden subscriptions or synthesising meeting notes into a Docs draft.

Daily Brief operates as an out-of-the-box morning digest that securely gathers urgent Gmail notes and Calendar meetings, helping you prioritise your day.

Also, there are “information agents” that would monitor the web 24/7 to send you synthesised updates on relevant topics like your favourite sports team or house hunting. Information agents will launch first for Google AI Pro & Ultra subscribers this summer.

Taiwo Kola-Ogunlade, Communications and PR Manager, Google West Africa, clarifies that these recent updates to Search do not mean Google is deprioritising traditional website links.

“We know that users want both AI and the web,” he explains. “We’re actually seeing that people are using AI to ask entirely new types of questions, which is creating new opportunities for content discovery. We are intentionally building our Search AI features to make it easy for people to connect to the web.”

He adds that Google just launched updates to how it shows links, including more inline links with website previews, new sections to help people explore in-depth articles, and links to first-hand advice from online discussions.

One of the recent updates is an intelligent hub that tracks items across Search, Gemini, YouTube, and Gmail, marking Google’s strongest push into “agentic commerce” and fundamentally altering how consumers will interact with online retailers, while streamlining checkout across brands.

The agents automatically track price drops, flag product hardware incompatibilities across different merchants, and utilise the new Agent Payments Protocol (AP2) to allow agents to securely make purchases on a user’s behalf within strict spending limits.

There’s also the introduction of live conversational voice controls and new image generation tools. With this, users can search Gmail, Docs, and Keep, hands-free, with the live voice offering conversational brainstorming sessions. Users can also execute precise object segmentation, altering specific items in an image without changing the rest of the picture in Google Pics.

With Ask YouTube, a conversational AI is natively integrated within the YouTube platform to summarise and interact with videos. Google Flow Music allows users to highlight and edit specific song sections: translating lyrics or changing a beat drop, without altering the rest of the original track.

Taiwo Kola-Ogunlade explains that agents taking action proactively on behalf of users is a major shift, so Google is starting with a controlled, phased release to ensure quality and safety.

“We are rolling out to the US in English first to gather user feedback on these complex new experiences, but we absolutely plan to expand to additional markets soon.”

Taiwo adds that feature availability will naturally vary by region and depend on the local regulatory landscape for AI services, and that Google always aims to make products available to as many people as possible.