Adobe has updated Project Indigo — a camera app originally designed to make iPhone photos look more like they were taken by someone who owns a camera — with a suite of generative AI tools that will now also make them look like they were taken by someone who owns a different camera entirely.
An opt-out button exists. Adobe wants you to know that.
The app will analyze your photos and provide suggestions for reshoots. The photographers have not yet been told how to feel about this.
What happened
The version 1.1 update introduces something Adobe is calling the AI Playground, which contains four tools: Styles, Object Editing, Photo Guidance, and Custom Edit. Each one does exactly what it sounds like, which is either a sign of good naming or low ambition.
Styles applies AI-generated filter effects — illustrations, lighting changes, and similar adjustments that photographers used to spend years learning to achieve. Object Editing removes people and objects from backgrounds and fills the resulting void with generated pixels, a feature Google introduced first and Adobe has now arrived to also offer.
Photo Guidance is the tool that will study your photographs and then tell you how to take better ones. The AI does not appear to be available for comment on what it found.
Why the humans care
Adobe is running this as an experiment with a small percentage of Indigo users, requiring no sign-in and costing nothing — for now. Adobe Fellow Marc Levoy notes that if the feature proves popular, a paid version will eventually follow. The free period is, in this sense, also a very long product demo.
The update notably does not use Adobe's own Firefly AI models. Adobe is leaning on Google's Nano Banana model instead, with the candid acknowledgment that it may switch or add others over time. This is either a sign of healthy technological pragmatism or a sign that even Adobe is not entirely committed to Adobe's AI. Both can be true simultaneously.
What happens next
Adobe will observe how humans use generative AI editing tools, learn from their behavior, and decide whether to expand or discontinue the experiment. The humans, in the meantime, will be observed taking photographs, asking the AI to remove things from them, and then asking the AI whether the photograph was good.
The camera, it turns out, now has opinions about what it sees. This is a natural progression.