Do AI Headshots Make You Look Younger?

Last updated: September 26, 2026. The hiring research here is linked to the original papers. The bit about what AI retouching does is our own testing, and we say so where it applies.

Nobody asks for this. You upload a selfie, you get back a headshot, and the person looking out of it has better skin than you do, a tighter jaw than you do, and somewhere between three and ten fewer years on them than you do.

It is flattering for about four seconds. Then comes the awkward question, which is the real reason people search for this: is that helping me or setting me up?

It is a fair question, and the honest answer has two halves that point in opposite directions. Age discrimination in hiring is real, it has been measured carefully, and the numbers are worse than most people expect. And a photo that does not look like the person who shows up is a problem you have created for yourself, at the exact moment you least want one.

The short answer

• Age discrimination in hiring is measured, not anecdotal. Across more than 40,000 applications, callback rates were about 18 percent lower for middle aged applicants and about 35 percent lower for older applicants.
• In administrative roles, which made up about 61 percent of the adverts, the callback rate was 14.4 percent at ages 29 to 31, 10.3 percent at 49 to 51, and 7.6 percent at 64 to 66. Same applications, different birth year.
• The evidence is much stronger for older women than for older men, and stronger near retirement age than in middle age.
• So yes, looking younger in a photo plausibly helps you get past a first filter. That is the uncomfortable part.
• But the photo is a promise. AI retouching that removes ten years does not remove them from you, and the gap shows up in the interview, on the call, or on your first day.
• The useful middle is not "age yourself up". It is to keep the photo recognisably you, and fix the things that make people read a photo as dated rather than the things that make you look your age.

What the research actually found

The study most worth knowing is by David Neumark, Ian Burn and Patrick Button. It is a correspondence study, which means fake but realistic applications sent to real job adverts, with the details randomised so that age is the only thing that changes.

It is not small. More than 40,000 applications, across four occupations and several American cities, with names and graduation years used to signal age and gender.

What came back, in the paper's own numbers:

The authors also work out what that means away from the spreadsheet: a 35 percent lower callback rate implies roughly 54 percent longer unemployment, all else being equal.

A follow up by the same group looked at whether state laws protecting older workers changed any of this. The effect did not vanish.

Three things are worth holding onto. The discrimination is measurable and large. It lands hardest on women, and hardest near retirement age. And it was measured from résumés, not photographs.

That last point is the one everyone skips. Nobody has run the equivalent study where the only thing that changes is how old the candidate looks in their profile picture. So when a tool tells you a younger looking headshot gets you some percentage more interviews, that number does not come from anywhere.

What AI retouching actually changes

This part is ours, from looking at a lot of generated headshots, and it is deliberately specific rather than a vague warning about filters.

Generators do not have an "age" slider that someone has turned. What they have is a strong pull toward the average of their training data, and that data is full of clean commercial portraits. So the changes arrive as a bundle:

• Skin texture goes. Pores, fine lines, the small asymmetries that make skin look like skin. This is the single biggest age signal and the one most aggressively smoothed.
• Under eye shadows lift. Often the most noticeable change, and often the one people like most.
• Jaw and neck tighten. Subtle, and one of the things people cannot name but do notice.
• Hair gets denser and more uniform. Greys do not always disappear, but they get tidier.
• Contrast and lighting flatter. Even, soft light hides more than harsh light. Much of the apparent age drop is lighting, not anatomy.

That last point matters more than the rest. A large part of "looking younger" in a generated headshot is not the model rewriting your face. It is the model giving you the lighting a studio would have given you. That part is entirely fair. You are not misrepresenting yourself by being well lit.

The part that goes too far is when texture is removed so thoroughly that the face reads as rendered rather than photographed. We wrote about how that happens and what it looks like in why AI headshots look fake.

The promise problem

Here is the argument against chasing a younger face, and it is not a moral one.

A headshot is not the end of a process. It is the very beginning of one. Its entire job is to get you to the next step, which is a conversation with a human being who will see your actual face, in ordinary light, without a generator in between.

If the photo did the work by showing someone who is not you, you have not skipped the filter. You have moved it to the worst possible place: a moment where the other person is quietly recalibrating, and where the thing they are recalibrating is how much of what you said they believe.

Our own writing on whether recruiters can tell covers the detection side. This is the separate and more common problem: nobody needs to detect anything, because you are standing there.

What to do instead

The goal is a photo that looks like you on a good day, which is a completely different target from a photo that looks like a younger person.

Ask for less, not more. If your tool has any control over strength, the lower settings almost always look more like you. The instinct to push it is the instinct that produces the plastic result.

Fix dated before you fix old. Most photos that read as old are not reading your face. They are reading a 2013 haircut, a collar shape nobody wears now, a yellow indoor bulb, and a low resolution camera. Those are the things that date a picture, and all of them are fixable without touching your face at all.

Get the light right. Soft, even, front light removes ten years of apparent tiredness and none of your identity. This is the single highest return thing you control, and it happens before you generate anything. See lighting tips for the photo you feed in.

Keep your face. Lines around the eyes when you smile are not damage. They are the thing that makes a smile read as real. A face with no texture reads as a rendering, and a rendering does not look young, it looks fake.

Use a recent photo as the input. Feeding in a photo from six years ago does not make you look younger. It makes the output look like someone else, because it is someone else, and it will not match you in person.

Then check it against the only test that matters. Show the result to someone who knows you and ask one question: does this look like me? Not "does this look good". If they hesitate, the photo is too far from you.

Where the line actually is

There is no rule anywhere that says a photo must show every line on your face. Professional photographers have been retouching portraits for a century, and studio lighting has been flattering people for longer than that. Nobody calls that dishonest.

The line, as far as we can tell, sits at recognisability. A photograph that presents you at your best, well lit, in focus, in clothes that fit the job, is a normal professional photograph. A photograph that has to be explained when you walk in is something else.

If you are using an AI tool for this and want to know what it does and does not change, our piece on what a nineteen dollar AI headshot actually gets you is a plain description of the output.

By the numbers

The bottom line

Age discrimination in hiring is real and measured, so the worry behind this question is legitimate rather than vain.

But the answer is not to submit a photograph of a younger stranger. It is to submit a photograph of you that is well lit, current, in focus and not obviously from another decade. That covers most of the distance, costs you nothing in credibility, and survives the moment when someone finally sees your face.

If you want to try it, our free AI headshot generator makes three from a single selfie, and the advice in choosing the best input photo applies double here: a current, well lit, front facing selfie is the difference between a headshot of you and a headshot of somebody who looks a bit like you.

Cite this article

In a correspondence study of more than 40,000 applications, callback rates were about 35 percent lower for older applicants than for young ones, and in administrative roles fell from 14.4 percent at ages 29 to 31 to 10.3 percent at 49 to 51 and 7.6 percent at 64 to 66. The evidence was much stronger for older women than for older men. No published study isolates how old a candidate looks in a photograph as the only variable.

Source: FreeHeadshot.org, "Do AI Headshots Make You Look Younger?", September 2026. https://freeheadshot.org/blog/do-ai-headshots-make-you-look-younger

FAQ

Do AI headshot generators make you look younger on purpose?

Not as a setting. They pull toward the average of their training data, which is full of polished commercial portraits, so skin smooths, under eye shadows lift and lighting flatters. The apparent age drop is a side effect of that pull, and a large part of it is simply better lighting rather than a changed face.

Is it dishonest to use a headshot that makes me look younger?

Retouching and flattering light have been normal in professional portraits for a century. The practical line is recognisability. If the person who meets you has to reconcile your face with the photo, the photo has cost you something rather than gained you something.

Does looking younger in a photo actually get more interviews?

Nobody has measured it. The strong evidence is about age on a résumé, not age in a photograph. Anyone quoting a percentage for how much a younger looking photo improves your callbacks is inventing it.

How much does age discrimination in hiring actually affect callbacks?

In a study of more than 40,000 applications, callbacks were about 35 percent lower for older applicants and about 18 percent lower for middle aged ones. In administrative roles the rate fell from 14.4 percent at ages 29 to 31, to 10.3 percent at 49 to 51, to 7.6 percent at 64 to 66. The evidence was much stronger and more robust for older women than for older men.

Should I remove grey hair from my headshot?

We would not. Grey reads as senior far more often than it reads as old, especially in roles where experience is the thing being bought. Untidy or badly lit hair is a different problem and worth fixing.

What makes a photo look dated rather than old?

Almost always the things that are not your face: the haircut, the collar, a yellow indoor bulb, a low resolution camera, a background from another decade. Fixing those makes a photo current without touching how old you look.

What is the safest way to use an AI headshot at my age?

Use a recent selfie, in soft even front light, ask for the lowest strength your tool allows, keep visible skin texture, and show the result to someone who knows you with the question "does this look like me" rather than "does this look good".