Gemini Spark Can Now Manage Your Google Photos

Google has connected Gemini Spark, its paid background agent, to Google Photos. Photos lead Shimrit Ben-Yair announced the integration in a post on X on 3 September, and it gives Spark standing permission to search, edit, organise and share a user’s photo library — either on request or on a schedule it then keeps to on its own.

The capabilities fall into four buckets. Spark can search and curate, finding photos and videos by subject, location, date or event, and it can be asked for the best shots from a set or told to filter duplicates. It can edit and create — enhancing quality, applying quick fixes, generating collages and stylised images — and every edit is saved as a new copy rather than written over the original. It can build albums, private or shared, and save curated selections into them, then push links into connected apps such as Gmail, Google Docs and Messages. And it can schedule automations, running recurring library tasks in the background.

Google’s own examples are the clearest guide to what it thinks the feature is for. One is a single compound instruction: find all my vacation selfies, erase the background crowds, and centre me in every shot before adding them to a new album. Another is a standing weekly job — pull the best shots of the kids, drop them into a shared album, and draft a recap email for the grandparents. A third turns a photographed concert flyer into a calendar appointment. Ben-Yair pitched it from her own library: “I’ve always dreamed of having a power agent to help me get the most out of my 143,206 photos and videos. And that day has come!”

Spark itself launched at Google I/O 2026 as a 24/7 personal agent built to run long, multi-step jobs in the background rather than answer questions in a chat window. Photos is a natural fit for that shape of agentic product: the work is tedious, well-defined and easy to verify by looking at the result, which is a far more forgiving setting than the ones where an agent’s mistakes are expensive.

The rollout is narrow for now. It reaches Google AI Pro and Ultra subscribers only, in the United States only, in English only, over the coming weeks; Google gave no timeline for other markets. Turning it on means connecting Google Photos to Gemini, toggling Spark in the app and then writing the prompt. Worth noting what that grant actually is: a background agent with continuing write access to a personal archive most people never curated in the first place. The decision to save every edit as a copy is the main guardrail on offer, and it is a sensible one — an agent that reorganises 143,000 photos is only useful if getting it wrong is survivable.