Google started rolling out Pics on September 1, and if you pay for AI Pro or your company runs a Workspace business plan, it's probably showing up in your account this week. Business Standard, Business Plus, Enterprise Standard, Enterprise Plus — all included. AI Pro for Education accounts are getting it as term starts.
The tool isn't new; Google showed these things off at I/O back in May. What's changed is where it lives.
It edits inside the document
Pics works on its own, but it also runs inside Docs and Slides, with Drive coming in a few weeks. That placement is the whole story. Editing an image used to mean leaving the document, opening something else, making the thing, exporting it, and dragging it back in. Pics collapses all of that into the file you were already working in.
It's built on Nano Banana, Google's image model, and the headline trick is object segmentation — grabbing one element in an image and changing it while everything around it stays put.
So: dropping a product onto a different backdrop, swapping a logo, resizing a piece of a chart, translating text that's baked into an image. Boring stuff. Also, for most people in most offices, the entire job — and the reason so many decks are full of screenshots that were never quite the right shape.
Who it's really for
The subscription gating gives away the target. This isn't for designers, who have better tools and good reasons to be wary of generated assets. It's for everyone else — the people who need a decent image, need it in slide four, and have neither the software nor the training to make one.
Google's also adding agentic video understanding to the AI Pro tier, which analyses video rather than generating it. Put that next to Veo for video and Flow for assembling the pieces and the strategy comes into focus: ship creative tools through subscriptions people already pay for, instead of selling a separate creative suite.
The part worth thinking about
Make something free and instant and you get more of it, including the stuff nobody needed. Anyone who's watched a document slowly fill with decorative stock photos knows how that ends.
There's a sharper version of this for businesses. Object segmentation makes small, invisible edits trivial — a changed label, a tweaked chart, a swapped background. Inside your own deck, that's efficiency. Applied to anything functioning as a record — a receipt, a screenshot, a photo someone treats as evidence — it's a different thing entirely. If this is landing across a whole department, it's worth deciding what the rules are before someone works it out on their own.
What to actually do
If it's arriving in your account, use it for what it's good at: resizing, backgrounds, on-image translation, logo swaps. Leave the from-scratch image generation to people who do that for a living.
The model isn't the interesting part. The interesting part is that the tool finally moved to where the work already was.
Image: George Milton, via Pexels
