Google is directly integrating AI image generation features into its "AI Overviews" in search. In the future, when users cannot find matching images on the web, they can simply enter a text prompt in the search bar to instantly generate the desired image.
This new image generation feature is based on Google's latest "Nano Banana2Lite" image model. The model is designed with speed and cost efficiency as top priorities, rather than striving for the highest image quality. It is reported that this feature currently only supports English users and will be gradually rolled out to all regions that support AI image generation in the coming weeks.
Industry analysts point out that this move is another step by Google to transform its search engine into an "AI-driven" platform aimed at retaining users. However, it may also further reduce traffic from the Open Web—although traditional image searches can still drive traffic to external websites, AI-generated results will clearly intercept this potential external traffic.
At the same time, "Google Images" is also about to receive a completely redesigned homepage. The new homepage will introduce a dynamic image library that captures online content in real-time and provides personalized recommendations based on each user's personal interests. Additionally, users will be able to save their favorite images to a collection, which will be displayed as tabs directly above the image library. The new homepage will also be rolled out in the coming weeks, starting with the English desktop version in the United States, and users will need to sign in to their Google accounts to access it.
