Phota Labs launches an iOS app that re-shoots everyday photos with AI

Phota Labs launched an iOS app on November 12 that it describes as a personal AI photo assistant. Instead of layering filters on an existing image, the app tries to reconstruct an ordinary snapshot into something that looks professionally shot, while preserving the way people actually look.

Founders Cecilia Zhang and Zach Xia met at Adobe, where they worked on computational photography products including Project Indigo, Lightroom and Photoshop. According to TechCrunch, they concluded that editing tools had become powerful but still could not rescue a photo that was badly captured in the first place.

The product's premise is to put the capture step back in with AI. A wide-angle group shot can be tightened into a closer portrait, and a selfie can be reframed as if someone else had taken it. Users pick a photographic style, ranging from subtle corrections that keep the original lens and composition to dramatic rewrites that introduce fisheye distortion, new angles or entirely different perspectives.

Unlike pointing a generative model at a photo and hoping for the best, Phota relies on its own personalization stack. The system learns the identity of the people in the image so it can keep their faces consistent through any transformation, a direct answer to the way generative tools tend to invent faces when a picture is blurry.

On the trust side, the app embeds an invisible watermark in edited images so platforms can identify them as AI-altered when they are shared. The founders frame the whole product around authenticity: the goal is a better photo of the same moment, not an AI-generated picture of a moment that never happened.

Commercially, Phota Labs raised a $5.6 million seed round led by a16z last year. The app is launching free to build an initial user base, with a freemium model to be explored later and an Android version on the roadmap.

The reason this is worth watching is that it targets a gap in consumer AI imaging: generative models are good at creating from scratch but weak at preserving real people, while conventional editors demand skill the average user does not have. If identity preservation and automatic reframing hold up at scale, the app could change how ordinary people work through a camera roll.