A key change is that Gemini 3.5 Flash now powers AI Mode in Google Search by default, making it the model behind the newest AI search features available to users worldwide.

The biggest upgrade to Google Search in 25+ years

Google also unveiled a major redesign of Search—described by the company as the largest upgrade to the search box in more than 25 years.

The new AI-powered interface allows users to ask conversational, complex queries rather than relying on short keyword searches. The system can also accept richer inputs such as images, files, or other content, allowing Search to better interpret user intent.

More importantly, the update introduces AI agents directly into Search. Instead of simply returning a list of links, users can now ask questions that trigger agents capable of gathering information and helping complete tasks.

Always‑on information agents

A central concept unveiled at I/O 2026 is the rise of Search agents—AI systems that operate continuously in the background.

For example, Google introduced information agents designed to monitor the web and other data sources for updates relevant to a user’s interests. These agents can track news, financial information, or product prices and deliver updates when something important changes.

This shifts Search from a reactive tool—something users open when they need information—into a system that actively watches for useful information and surfaces it automatically.

Universal Cart: AI‑managed shopping across Google

Google also announced Universal Cart, a Gemini-powered shopping experience designed to unify purchasing activity across different Google services.

The system can collect products that users encounter across Search, YouTube, Gmail, and the Gemini app, organizing them into a persistent shopping cart that continues across sessions. It can also monitor prices and deals to help users decide when to buy.

Instead of starting a new shopping search each time, the idea is that Gemini acts as a shopping assistant that tracks products, compares offers, and helps complete purchases over time.

Android XR smart glasses

Another major reveal was Android XR smart glasses, part of Google’s new extended‑reality platform designed for headsets and eyewear.

The glasses integrate Gemini directly into everyday situations, allowing users to interact through voice and receive contextual assistance such as directions, translations, and notifications without pulling out a phone.

Google also previewed partnerships with eyewear brands Warby Parker and Gentle Monster to design consumer‑ready frames, with early versions expected to launch in the future.

Gemini Omni: A new multimodal AI model

Alongside the Gemini 3.5 series, Google introduced Gemini Omni, a new multimodal model designed to generate content from mixed inputs such as images, audio, video, and text.

The first version of Omni focuses on video generation, allowing users to combine multiple types of input and create or edit video through conversational prompts.

The bigger strategy: Search as an “AI agent manager”

Taken together, the announcements reveal Google’s broader strategic shift.

For decades, Google Search worked primarily as a link discovery system—users typed keywords, and Google ranked pages across the web. The new AI features aim to transform it into something closer to a task manager powered by AI agents.

In this model:

• Users express goals or questions in natural language
• Gemini interprets the intent behind the request
• AI agents gather information, monitor changes, or perform actions
• Google services help complete the task

By embedding Gemini across Search, shopping, apps, and wearable devices, Google is positioning AI not just as a chatbot—but as the operating layer for how people interact with the internet.

Google I/O 2026 made that vision explicit: the future of Search may be less about finding links—and more about delegating work to intelligent agents that operate across the entire Google ecosystem.

A key change is that Gemini 3.5 Flash now powers AI Mode in Google Search by default, making it the model behind the newest AI search features available to users worldwide.

The biggest upgrade to Google Search in 25+ years

Google also unveiled a major redesign of Search—described by the company as the largest upgrade to the search box in more than 25 years.

The new AI-powered interface allows users to ask conversational, complex queries rather than relying on short keyword searches. The system can also accept richer inputs such as images, files, or other content, allowing Search to better interpret user intent.

More importantly, the update introduces AI agents directly into Search. Instead of simply returning a list of links, users can now ask questions that trigger agents capable of gathering information and helping complete tasks.

Always‑on information agents

A central concept unveiled at I/O 2026 is the rise of Search agents—AI systems that operate continuously in the background.

For example, Google introduced information agents designed to monitor the web and other data sources for updates relevant to a user’s interests. These agents can track news, financial information, or product prices and deliver updates when something important changes.

This shifts Search from a reactive tool—something users open when they need information—into a system that actively watches for useful information and surfaces it automatically.

Universal Cart: AI‑managed shopping across Google

Google also announced Universal Cart, a Gemini-powered shopping experience designed to unify purchasing activity across different Google services.

The system can collect products that users encounter across Search, YouTube, Gmail, and the Gemini app, organizing them into a persistent shopping cart that continues across sessions. It can also monitor prices and deals to help users decide when to buy.

Instead of starting a new shopping search each time, the idea is that Gemini acts as a shopping assistant that tracks products, compares offers, and helps complete purchases over time.

Android XR smart glasses

Another major reveal was Android XR smart glasses, part of Google’s new extended‑reality platform designed for headsets and eyewear.

The glasses integrate Gemini directly into everyday situations, allowing users to interact through voice and receive contextual assistance such as directions, translations, and notifications without pulling out a phone.

Google also previewed partnerships with eyewear brands Warby Parker and Gentle Monster to design consumer‑ready frames, with early versions expected to launch in the future.

Gemini Omni: A new multimodal AI model

Alongside the Gemini 3.5 series, Google introduced Gemini Omni, a new multimodal model designed to generate content from mixed inputs such as images, audio, video, and text.

The first version of Omni focuses on video generation, allowing users to combine multiple types of input and create or edit video through conversational prompts.

The bigger strategy: Search as an “AI agent manager”

Taken together, the announcements reveal Google’s broader strategic shift.

For decades, Google Search worked primarily as a link discovery system—users typed keywords, and Google ranked pages across the web. The new AI features aim to transform it into something closer to a task manager powered by AI agents.

In this model:

• Users express goals or questions in natural language
• Gemini interprets the intent behind the request
• AI agents gather information, monitor changes, or perform actions
• Google services help complete the task

By embedding Gemini across Search, shopping, apps, and wearable devices, Google is positioning AI not just as a chatbot—but as the operating layer for how people interact with the internet.

Google I/O 2026 made that vision explicit: the future of Search may be less about finding links—and more about delegating work to intelligent agents that operate across the entire Google ecosystem.