Adobe’s ‘natural look’ camera app embraces generative AI
Adobe is testing generative AI in Project Indigo, adding filters, object removal, and edit guidance to its iPhone camera app.
Intelligence analysis by GPT-5.4 Mini
Project Indigo started as a camera app for more natural-looking iPhone photos, but Adobe is now layering in generative AI through an experimental AI Playground. The move widens the app from capture tool to capture-and-edit platform, while also raising familiar questions about transparency and trust.
Adobe turned a camera app that tried to make photos look natural into one that can also change pictures with AI, like handing a painter a magic eraser and a costume box. It is testing the idea on only a small group first.
Analysis
From Camera Nerd Tool To AI Sandbox
Project Indigo launched as a deliberate counterpoint to heavy-handed mobile image processing, aiming for a more natural, SLR-like result. That makes this update notable: Adobe is not just adding another editing feature, it is redefining what the app is for.
The tension is obvious. A product built around photographic realism now includes generative styles, object removal, and broader prompting tools that can materially change what the image depicts. Adobe seems to be betting that users want both trustable capture and flexible transformation in the same workflow.
Adobe Is Treating This Like A Live Experiment
Adobe says the new AI Playground is an experiment, and that only a small percentage of Indigo users will see it over the next few weeks. The company is also letting people opt out and keep using the app as before, which signals caution rather than a full product pivot.
That framing matters because it shows Adobe is still learning where users draw the line. Marc Levoy’s comments suggest the company wants to measure whether people actually use GenAI-powered editing before deciding whether to expand access, add more experiments, or eventually charge for it.
The Real Test Is Trust, Not The Feature List
The feature set itself is straightforward: styles that generate new looks, object editing that fills in removed areas, photo guidance that suggests reshoots, and custom edits that respond to detailed prompts. Adobe is even leaning on Google’s Nano Banana model to start, which underlines how dependent this kind of mobile AI can be on outside model providers.
The harder problem is credibility. Adobe says it is working toward C2PA Content Credentials for manipulated images, while Google’s model already uses an invisible SynthID watermark, but the article makes clear that transparency is still catching up to capability. If Indigo is going to blur the line between camera and generator, it will have to prove that users can still understand what they are sharing and how they made it.
Key points
- Project Indigo is adding an experimental AI Playground with generative editing features.
- The new tools include Styles, Object Editing, Photo Guidance, and Custom Edit.
- Adobe says users can opt out and keep using the app the old way.
- The company is testing the feature with a small group and may later expand it or charge for it.
- Adobe says it is working toward C2PA Content Credentials for AI-edited images.
If the experiment works, Indigo could become a rare app that combines careful photo capture with powerful editing in one place. That could make advanced image tweaks easier for regular phone users without forcing them into separate tools.
The update could also weaken the app’s original appeal if users feel the “natural look” promise is getting diluted. Even with watermarks and content credentials, AI edits can still make it harder for people to trust what they are seeing.
