Google has moved its AI workspace directly into Search for all US users, putting document creation, coding, and app building one click away from billions of daily queries. The Canvas feature inside AI Mode now reaches every English-speaking user in the United States after eight months of limited testing through Google Labs.
Unlike ChatGPT's automatic workspace activation, Google requires users to deliberately select Canvas from the tool menu while in AI Mode, reflecting a design philosophy that prioritizes intentional interaction over automated assistance.
Canvas opens a persistent side panel where users can draft documents, refine creative writing, or generate code for shareable apps without leaving their search session. The workspace pulls live information from web searches and Google's Knowledge Graph directly into working documents, grounding outputs in current data rather than static model knowledge.
Coding capabilities arrive with this full rollout after being absent from the initial Labs trial period. Users can describe an idea to Canvas and watch as it generates functional code for apps or games powered by Gemini 3, requiring no development environment setup.
Once built, prototypes can be tested and refined through direct conversation with Gemini within the same panel. The feature represents a shift for Google, embedding creation tools into Search rather than confining them to separate applications like Gemini or NotebookLM. This positions Canvas to reach tens of millions of users who have never visited Google Labs or subscribed to premium AI services.
Canvas first appeared as an experimental feature in July 2025 alongside live video analysis and PDF uploads during an AI Mode upgrade. The eight-month testing period allowed Google to expand functionality beyond simple travel planning visualizations to include full document drafting and app generation capabilities.
Google's approach differs from competitors in both timing and design philosophy. While ChatGPT rolled out its Canvas feature to all users in December 2024 with automatic activation based on query analysis, Google waited until March 4, 2026 for a broad release that requires manual user initiation through the tool menu.
The workspace sits alongside search results rather than replacing them, allowing users to continue querying while feeding new findings directly into their documents without switching tabs or losing context. This integration makes drafts and interactive tools possible from inside Search itself rather than requiring separate applications.
Canvas also exists within the standalone Gemini app where Pro and Ultra subscribers access Gemini 3 with its 1 million-token context window for larger projects. However, the Search integration targets mainstream users who may never explore dedicated AI applications.
Google has already signaled confidence in this direction by offering US college students a free year of Google AI Pro access earlier this year, with signup available until January 31, 2026.
