Ask Runable forDesign-Driven General AI AgentTry Runable For Free
Runable
Back to Blog
Technology7 min read

Buckle up: Google is set to remake search with agentic AI in 2026 - Ars Technica

Google's AI search evolution is accelerating at I/O 2026. Discover insights about buckle up: google is set to remake search with agentic ai in 2026 - ars techni

TechnologyInnovationBest PracticesGuideTutorial
Buckle up: Google is set to remake search with agentic AI in 2026 - Ars Technica
Listen to Article
0:00
0:00
0:00

Buckle up: Google is set to remake search with agentic AI in 2026 - Ars Technica

Overview

Buckle up: Google is set to remake search with agentic AI in 2026

Google’s AI search evolution is accelerating at I/O 2026.

Details

Last year marked the beginning of Google’s explicit focus on AI search, and this year’s I/O solidified that shift. As Google’s search VP Liz Reid said during the keynote, “Google search is AI search.” This change is well underway, and the very reasonable objections to this path will not dissuade the company. All the metrics that matter to Google say this is the right move. But at the end of the day, Google can get whatever outcome it wants because it’s just that big and influential.

Google started testing AI Mode for search just over a year ago, making the shift official at I/O 2025. You hear a lot of complaints around the Internet about how AI is changing Google’s search products, but Google is getting what it wants: more searches. Reid revealed at I/O 2026 that AI Mode usage has been doubling every quarter. There are now more than 1 billion people using AI Mode every month.

It’s not hard to see how that could be true. AI Mode invites a conversational experience—it asks you questions—and each of those follow-up queries counts as searches. Google has also pushed AI Mode very hard, including prominent links and nudges to get people to use its search chatbot instead of the traditional product. And unlike many of Google’s other AI experiences, you don’t have to pay anything to AI search. Everyone who uses Google search gets the full AI experience.

You can hardly escape AI Mode as it is, and Google is announcing even more AI Mode integrations at I/O this year. AI Overviews may be the most prominent element of Google’s AI search shift, but that’s increasingly looking like a stopgap as AI Mode spins up. Google has a new “seamless” search experience that ties AI Mode into AI Overviews. Most Google searches now produce an AI Overview. Google is expanding a mobile feature that lets you move from an Overview into AI Mode. This feature is now available across desktop, too.

The AI Mode nudge hovers at the bottom of the Overview, actually hiding the top of the organic search results. This will, no doubt, inflate the number of AI Mode searches even more. It may also disincentivise users from scrolling down to see the 10 blue links. It makes organic results feel more like footnotes than the core of the search experience.

Reid also says Google’s new search box is the biggest change in its entire 25-year history. A search box isn’t very complicated (or didn’t used to be), but the new version again guides users toward AI Mode. It expands dynamically as you type more, and it will attempt to autocomplete your searches. Google definitely doesn’t want you to call it autocomplete, though! It uses generative AI technology to guess at your intent, guided by what Gemini knows about you. This change is rolling out today globally.

Google’s more-AI-than-ever search experience is also veering into experiences that don’t really feel like a search engine. Search will employ agents to answer your questions in new ways, powered by Gemini 3.5 Flash. Google says it has integrated Antigravity as a harness for the new model’s AI agents in search. This powers two different ways of finding information with vibe coding that are similar but technically distinct.

When you ask questions in regular search (AI Overviews) or AI Mode later this summer, Google’s AI may choose to create generative UI. These are single-shot simulations that help you understand concepts like, for example, the golden ratio or the behavior of black holes. These interfaces will have sliders, buttons, and other elements conjured from the AI ether.

AI Mode builds custom mini-apps from a single prompt, but it might not show you the code at launch.

The other experience is currently limited to AI Mode, and it goes a step further. If your query calls for it, Search will create a custom app to help you with the problem. Currently you have to ask for an app (e.g. make or build ‘x’ for me), but the line between generative UI and apps may blur over time.

What is that supposed to do for you? Maybe you want to plan a family outing for the weekend, so you ask search to build an itinerary. In that case, Search can create a UI with event suggestions, reviews, map embeds, and calendar integration. It pulls this data from Google’s platform as well as from around the web. The early demos of search agent dashboards actually show you the code as it’s generated, but Google is most likely going to hide this for the full rollout later this summer. Showing a simplified workflow of chain of thought would avoid confusing the average user who just wants a pretty UI and doesn’t care that it was generated on the spot.

You can revisit and change the dashboard by accessing your AI Mode history in the sidebar. These generated apps can be customized with follow-up prompts, and you can share them with others via a link. The other party can even customize the app to their liking. Currently, there is no way to share those modifications, but that’s something Google is exploring. It may also be possible in the future to manually modify the code of these mini-apps line by line.

The overarching trend here is fewer blue links and more AI-generated everything. Google says the greater efficiency of Gemini 3.5 Flash enables all these new AI experiences, and we can expect more of them in the future. The agentic app generation in particular may benefit from the pending improvements in Gemini 3.5 Pro, which might even be available before everyone gets search agents.

Googlers talk about these moves as a way to more efficiently extract the information people want from webpages that have become weighed down with extraneous text that forces you to scroll past more and more ads. That is a genuine problem with the current state of web content, but Google’s hands aren’t clean. Many websites have ended up in this state only after years of chasing search rank and compensating for low ad rates.

Despite what many see as a decline in Google search quality, the company’s search products remain far and away the primary way people find things online. Even after a year of Google’s AI search overhaul, Duck Duck Go, Bing, Brave, and the rest of the competition continue to be little more than a rounding error. Google appears to take its continued dominance and growth as proof that it’s on the right track with AI.

Google has decided this is how search works, and the rest of us are just along for the ride.

  1.          Plex's 200% Lifetime Pass price hike tries forcing users to another subscription
    
  2.          Iran demands Big Tech pay fees for undersea Internet cables in Strait of Hormuz
    
  3.          Civilization VII finally lets you build a civ that stands the test of time
    
  4.          Spider-Noir final trailer gives us a classic villain
    
  5.          The Internet can't stop watching Figure AI's humanoid robots handling packages
    

Ars Technica has been separating the signal from the noise for over 25 years. With our unique combination of technical savvy and wide-ranging interest in the technological arts and sciences, Ars is the trusted source in a sea of information. After all, you don’t need to know everything, only what’s important.

Key Takeaways

  • Buckle up: Google is set to remake search with agentic AI in 2026

  • Google’s AI search evolution is accelerating at I/O 2026

  • Last year marked the beginning of Google’s explicit focus on AI search, and this year’s I/O solidified that shift

  • Google started testing AI Mode for search just over a year ago, making the shift official at I/O 2025

  • It’s not hard to see how that could be true

Cut Costs with Runable

Cost savings are based on average monthly price per user for each app.

Which apps do you use?

Apps to replace

ChatGPTChatGPT
$20 / month
LovableLovable
$25 / month
Gamma AIGamma AI
$25 / month
HiggsFieldHiggsField
$49 / month
Leonardo AILeonardo AI
$12 / month
TOTAL$131 / month

Runable price = $9 / month

Saves $122 / month

Runable can save upto $1464 per year compared to the non-enterprise price of your apps.