Google has introduced an AI-powered image generator feature within its Search Generative Experience (SGE), initially available to users in the United States. This feature allows users to initiate image generation by using phrases like “draw” or “sketch” in the search bar or by exploring image search results that may trigger image creation prompts.
The “Similar Items” feature in Google Search allows users to create up to four new images simultaneously. These generated images are displayed above regular search results with a distinct background color. Users can open a generated image in a side panel, where Google records the specific prompt or description used, indicating whether it’s photorealistic.
Once a user selects an image from the search results, they can export it as a PNG or save it to Google Drive. The “Edit” option directs users to a dedicated page for refining the search prompt. A 30-day history of recent images is accessible on google.com/search/images.
Google has implemented measures to block images depicting harmful, misleading, or explicit content. Prompts involving notable individuals will not be generated, and SGE will not create photorealistic faces. Every image produced by SGE will have metadata labels and embedded/invisible watermarking to denote its AI-generated nature. Daily limits on image generation have been established, with specific details not disclosed.
The image generator is initially launched for US English users, with a broader rollout planned for the future. Access to this feature requires enrollment in SGE via Google Labs and user age of 18 or above.
