A new free iPhone app called Phota promises to do for ordinary snapshots what a professional photographer and a long editing session used to do, turning a hurried selfie or a wide group photograph into a polished portrait while keeping the people in it looking like themselves. The app, from a startup called Phota Labs, picks photographs from a user's camera roll and reworks them automatically, aiming to fix the framing, the angle and the light without changing the face staring back from the screen.
The idea is aimed squarely at the millions of photographs that are taken every day and never shared because something about them feels slightly wrong. The light was poor, the angle was unflattering, someone blinked, or the shot was taken at arm's length and looks like a selfie rather than a portrait. Most people do not have the time or the skill to correct those faults by hand, and general purpose AI image tools often make things worse by subtly changing a person's appearance, particularly when the original photograph is blurry. Phota Labs says it built its own technology to learn what a person actually looks like from their photographs and to hold on to that likeness through every edit.
How Phota Edits Photos
Phota can take a wide-angle group shot and tighten it into a portrait, or make a selfie look as though someone else was holding the camera. Users can choose styles that keep the original lens and composition, or they can ask for bigger changes, such as a fisheye effect, a different angle or an entirely new perspective. The app needs no written prompt. It works out what to change on its own, and according to the company its tests show that people prefer its results to those of general purpose tools in nine out of ten side-by-side comparisons.
The emphasis on keeping people recognisable is the heart of the product. Co-founder Cecilia Zhang told the technology publication TechCrunch that tools such as ChatGPT can alter a person's appearance when they rework a photograph, which defeats the purpose of a picture meant to record a real moment. Phota's answer is a system trained specifically on photography, designed to correct closed eyes, awkward expressions and poor lighting while keeping identity and context intact. Edited images also carry an invisible watermark that identifies them as AI-altered when they are shared on social platforms, a quiet acknowledgement of the growing concern about realistic images circulating without any sign of how they were made.
From Adobe to Phota Labs
Zhang and her co-founder Zach Xia met at Adobe, where they worked on computational photography products including Project Indigo, Lightroom and Photoshop. Both are among Adobe's leading artificial intelligence researchers, and the pair saw room for an app that could rescue a poorly captured photograph and handle the editing without asking the user to learn anything. Xia told TechCrunch that some of the app's beta users began taking more photographs simply because they knew the app could improve them afterwards, a habit that suggests the product may change how casually people shoot as much as how their pictures look.
The startup is backed by serious money. Phota Labs raised 5.6 million dollars in a seed round led by Andreessen Horowitz, the venture capital firm known as a16z, in 2025, and the firm announced its investment in a post in September of that year. The same technology sits behind a second product, an application programming interface for developers who want to build identity-preserving image tools of their own, including the ability to place the same person into different scenes or swap a character into a photograph without losing the likeness. That developer side of the business may matter as much as the consumer app in the long run.
Pricing and Android Plans
Phota is free on iPhone at launch. The startup plans to build an initial user base first and explore a freemium model later, charging for advanced features while keeping the basic experience free. An Android version is also planned, though no date has been given. The company is initially targeting people who take and share a lot of photographs and would rather skip manual editing altogether, and its founders also see a use for parents assembling family albums and grandparents preparing photographs to send to their grandchildren.
Whether the app finds a lasting audience will depend on trust as much as on technology. AI photo tools have multiplied quickly, and many of them produce flattering images that look nothing like the person who posed for them. Phota's bet is that people want better photographs of the moments they actually lived, not prettier versions of someone else. If its identity-preserving approach works as advertised, it could sit in the space between a camera filter and a full editing suite, doing in seconds what once took an afternoon. For now the only way to judge that claim is to try it, which is presumably why the app costs nothing to download.
