Google has announced a new set of AI modes in searches, significantly improving the way users interact with artificial intelligence tools. New updates include support for live video queries via Google Lens, the ability to analyze PDF files, and dynamic canvas mode designed for planning and organizations.
In a blog post shared this week, Google is expected to gradually unfold four key features, with many people expecting to reach US users first. These updates will more closely align AI modes with Gemini, Google's wider AI assistant, by offering multimodal input options and deeper integration with Chrome and Drive.
The most notable addition is the search live extension. First launched with real-time audio interaction, this feature now supports live video via Google Lens. Users can open the Google app and tap the live icon in the lens to ask questions about what will be displayed in real time. The AI mode analyzes the video feed and responds accordingly. This feature will begin rolling out on mobile devices in the US this week for users registered in the AI Mode Lab experiment.
Google has also added document analysis capabilities to its AI mode. Starting next week, desktop users will be able to upload PDFs and ask questions about their content. Chatbots contextualize your data and provide answers and insights. Google has also confirmed that support for additional file types and integration with Google Drive will continue.
A new feature called Canvas has also been introduced. This tool allows users to create detailed plans, such as travel itineraries and study schedules, directly within AI mode. The canvas does not have a standalone button, but the associated query triggers the Create Canvas prompt. Upon launching, the tool opens a side panel where users can improve or modify plans through follow-up questions. We will soon be able to support personalization using files such as class notes and syllabus.
Another update will provide AI mode directly to the Chrome browser. By selecting “Ask Google About this Page” from the address bar, users activate a side panel where the AI assistant can answer questions about any element that appears on the screen. This allows users to interact with web content contextually, further blurring the line between search and live digital support.
