Adobe's 'camera nerd' app pivots hard into Google-powered AI

The Indigo camera app, Adobe’s experiment in giving iPhone photos an SLR-like look, just took a hard left turn into generative AI. And the company isn’t even using its own models to do it. Adobe is updating Indigo with an ‘AI Playground’ suite that leans on Google’s Nano Banana model—at least for now. Marc Levoy, an Adobe fellow, framed the move as a learning experiment. ‘Our goal is to learn how people use GenAI-powered editing,’ he said.

The update, rolling out to a small percentage of users over the next few weeks with no sign-on required, splits its new tricks across four sections: Styles, Object Editing, Photo Guidance, and Custom Edit. The Object Editing tool works much like Google’s Magic Eraser, removing people and distractions while generative fill patches the holes. Custom Edit lets you describe broader changes in natural language if the preset styles miss the mark. Levoy noted that hitting a button again will ‘typically get something slightly different’ if you don’t like the first result.

This is a sharp departure from Indigo’s launch identity. My colleague described it then as a ‘camera app built by camera nerds for camera nerds,’ focused on manual controls and computational photography to mimic big-glass depth of field. Adobe may have seen this coming, though. Levoy had previously described Indigo as the start of a journey toward an integrated camera and AI editing experience. What’s less expected is the lack of Adobe’s Firefly models in the mix, and the somewhat looser approach to content transparency.

Levoy says the team is working on adding C2PA Content Credentials to manipulated images, and points out that Google’s SynthID watermark is already applied by Nano Banana. Still, the ethical hand-wringing feels a step behind Adobe’s usual posture. ‘In the end, nothing can be implemented that reduces the importance of behaving responsibly,’ Levoy said. If the feature gains traction, a paid version is on the roadmap. For now, it’s a free experiment that asks an interesting question: will the same users who wanted an SLR aesthetic also want to turn their photos into AI-generated illustrations?