Mountain View, California, April 19, 2026, 10:31 PDT
Google is rolling out a Gemini update that lets the chatbot use a subscriber’s Google Photos library to create personalized AI images, pushing the company’s consumer AI deeper into private account context rather than relying only on typed prompts.
The feature links Personal Intelligence, Google’s opt-in system for connecting Gemini to Google apps, with Nano Banana 2, its image-generation model. It can use Google Photos labels to identify people such as the user, friends and family, then generate an image from a simpler request, The Verge reported after Google’s announcement.
The timing matters because Google is trying to make personal data an advantage in the paid AI race. Image generation has become a front door for consumer AI subscriptions, and Google is using a deep Photos library, not just a smarter prompt box, to cut the work needed to make tailored pictures.
Google said the update is rolling out over the next few days to eligible U.S. subscribers on Google AI Plus, Pro and Ultra, with plans to bring it to Gemini in Chrome on desktop and to more users later. TechCrunch said the tool can also show a “sources” button so users can see how Gemini derived context for a generated image. TechCrunch
Google product managers Animish Sivaramakrishnan and David Sharon wrote that the change is meant to let people spend “more time creating and less time explaining.” Their examples include asking Gemini to design a dream house or create a picture of desert-island essentials, with the output reflecting tastes from connected Google apps. Blog
In practical terms, the update means a user no longer needs to search for a photo, download it and upload it again to make a family-style image. If Photos labels are set, Gemini can use a relevant image as a reference; users can correct the result or choose another reference photo if the model picks the wrong detail.
The risk is trust. Google says Gemini does not directly train its models on a private Google Photos library and that app connections can be turned off, but a Google Photos help page says Gemini features in Photos can use photos, videos, face-group labels and Google Account information, and may process photos and videos to improve edits or make inferences, including guesses about ages and locations of top face groups.
Early hands-on testing suggests the feature can preserve a user’s likeness across varied scenes. TechRadar’s Eric Hal Schwartz wrote after trying it that a fantasy image kept his face recognizable rather than using “a generic placeholder,” though he also noted some outputs drew odd inferences. TechRadar
That puts Google against OpenAI’s ChatGPT Images, Adobe Firefly and Meta AI, all of which market image generation or editing from text and reference images. The difference Google is pressing here is first-party context: the tool can draw from Photos and other connected apps when the user opts in.
Personal Intelligence itself is not new. Google introduced it in January as a way for Gemini to answer personal requests by referencing apps including Gmail and Photos; the company said then that app connections are off by default and users can adjust settings, disconnect apps or delete chat history.
The open question is how many paid users will connect Photos for image creation, and whether the source controls will be clear enough for normal users. If the tool feels intrusive, or if it pulls the wrong family member or private memory into a prompt, the convenience case weakens fast.
For Google, the bet is that a personalized image model will make Gemini feel less like a generic chatbot and more like a service built around a user’s life. For users, the choice is narrower and more immediate: better images, or less account context in an AI system.
