Google Images pivots to discovery model with Pinterest-style overhaul and AI generation
The search engine’s image platform is moving away from a traditional search-centric interface to a dynamic, personalised feed, while also integrating AI image generation capabilities into its AI Overviews feature.
Google is overhauling its Images homepage to coincide with the platform’s 25th anniversary, shifting from a traditional search-centric interface to a discovery-based model. The redesign introduces a dynamic, Pinterest-style feed for signed-in desktop users in the United States, featuring a real-time gallery tailored to individual interests and browsing history. This update marks a significant departure from the previous interface, which consisted largely of a blank page with a search bar, replacing it with a dense, scrollable stream of visual content.
The new homepage operates as a “browseable” feed that continuously updates with images from across the web as users scroll. To facilitate organisation, Google has introduced “collections,” allowing users to save specific ideas into dedicated tabs above the main gallery. These collections are intended for practical inspiration, enabling users to curate and reference visual content without initiating a specific search query. The rollout is currently limited to English-language desktop users in the US, with broader expansion planned for the coming weeks.
Concurrently, Google Search is integrating AI image generation within AI Overviews, utilising the Nano Banana 2 Lite model to assist with visualisation tasks. Demos provided by the company illustrate how the feature can be used for comparing or visualising home decor ideas. The system responds to prompts such as “help me visualize” or “create a visual” to generate relevant imagery, aiming to streamline the process of conceptualising design or aesthetic choices.
The deployment of these features follows a phased strategy. The visual redesign is being deployed on desktop devices in the United States for English-language users with a signed-in Google Account. The AI image generation component will follow a similar trajectory, expanding over the coming weeks to all regions that currently support image creation in AI Mode. This staged approach allows Google to manage the technical demands and user feedback associated with the new functionalities.
However, details regarding the operational safeguards of the AI features remain sparse. Google has not yet disclosed the specific criteria that trigger image generation in AI Overviews, nor has it provided information on how it prevents the system from generating images in sensitive contexts, such as current events. The full timeline for the global rollout of the AI image generation feature remains unspecified beyond the general timeframe of “coming weeks.”
