Why Flawless Retouching Now Looks Like AI

For about a decade, glass-smooth skin and spotless frames marked a careful retoucher, and that exact look is now what image generators produce best. If a client scrolling your portfolio can't tell your work from a prompt, they have one less reason to book you.

Prince Meyson has been shooting for over 16 years, and he says he has never been less sure where photography is heading. He also admits he helped create the look he's now warning against. Early in his career, he smoothed skin hard and pushed every image toward perfection, and models trained on years of that kind of work can now produce it faster than any person can. His first fix is the simplest: stop overediting. AI retouching is built into Photoshop and dedicated tools like Evoto. Trouble starts when a few sliders make it too easy to go too far. Meyson still uses those tools. He argues that freckles, moles, and small flaws now carry more weight than ever, since they are some of the clearest signs that a real person stood in front of your lens.

Generators learn from what is already online. The more your concept looks like a Pinterest board, the more it looks like something a model could produce. Meyson is clear that he isn't asking for something nobody has ever seen. Taking a reference and putting your own spin on it is enough to pull an image away from the averaged look of everything a model has absorbed. He applies the same thinking to generative tools inside editing software. Removing a stray bracelet with generative fill is a sensible shortcut. Building a new background, stretching a dress into yards of flowing fabric, and placing your subject somewhere they never stood is where an image starts to read as synthetic.

In 2024, Meta began tagging posts on Instagram and Facebook as "Made with AI." Working photographers quickly found that label on real images where they had done little more than a small generative cleanup in Photoshop. After the backlash, Meta changed the wording to "AI info" and, for images that were only edited, moved it into a menu instead of showing it on the post. Around the same time, camera makers started building proof of origin into the hardware. The Leica M11-P launched in 2023 as the first camera to embed Content Credentials, a signed record of how and when an image was captured. Sony has since added similar C2PA signing to some of its professional bodies through firmware. Those systems are still uneven, and many platforms strip or ignore the data before anyone sees it. Until that changes, a client judging your work mostly has two things to go on: the image itself and whatever you can show about how it was made.

Meyson's answer to heavy generative work is to bring more of the final image into the studio. That means real props, real fabric, and real light, so that post-production becomes cleanup instead of construction. He points to big-budget filmmakers returning to practical effects as the same instinct at a larger scale. "Get it right in camera" is old advice, and he says it means more now than at any point in his career. He also acknowledges how far the other side has come. A prompt can now ask for a natural edit with a few imperfections left in, so restrained retouching on its own proves less than it did even a year or two ago. The first of the two habits he calls most important is showing your behind-the-scenes process, in video or stills, so people can see how you lit, propped, and shot an image.

The second habit Meyson considers most important is in the video above.