You snapped a photo, uploaded it to ChatGPT, and got your answer. But what if that answer is wrong more than half the time?
In the shower one morning, you notice a spot.
It doesn’t look like your other moles. Wrong color. Weird shape.
Your stomach drops.
Is it cancer? Part of you doesn’t want to know the answer to that question.
So you don’t call a doctor. Not yet.
Instead, you open Google.
Googling “what does skin cancer look like” doesn’t help much.
The images are confusing. The descriptions are vague. You’re more anxious than when you started.
So you snap a photo of the spot and upload it to ChatGPT. You type: “Is this melanoma?”
ChatGPT says no. It’s probably nothing.
You exhale. Close the app. And move on with your day.
But here’s the problem: ChatGPT was probably wrong.
What Does It Actually Take to Diagnose Skin Cancer?
Your dermatologist doesn’t just look at a photo.
They look at your actual skin. In person. With a trained eye that’s examined thousands of spots.
They might use a dermatoscope, a small handheld device that uses light and magnification to see details you’d never catch with a camera.
It doesn’t hurt. It takes seconds. And it reveals things your smartphone photo simply can’t.
If something looks suspicious, a biopsy follows. That tissue goes to a lab. The result, combined with what the doctor saw in the room, is what becomes a diagnosis.
A photo can’t do any of that.
And that’s before we consider that most photos people take of their own skin are blurry, poorly lit, or shot at the wrong angle.
ChatGPT is working with bad data before it even starts.
So How Accurate Is AI at Diagnosing a Concerning Spot From a Photo?
According to a 2026 study published in the Journal of the American Academy of Dermatology, not very.
ChatGPT accurately diagnosed melanoma, the deadliest form of skin cancer, only 47% of the time.
Let that sink in. Worse than a coin flip.
But it gets scarier.
When researchers uploaded images of actual skin cancer to ChatGPT, it read the spot as benign (nothing to worry about) 51% of the time.
More than half the time, it gave a false sense of security to someone who needed to be seen.
The average dermatologist, by comparison, accurately diagnoses suspicious spots 70% of the time, based on the same research.
And Dr. Yentzer at Finger Lakes Dermatology? His accuracy is likely higher than the national average.
That gap, between what AI can do and what a trained dermatologist can do, isn’t a minor technical footnote.
It’s the difference between catching something early and missing it entirely.
Why Does It Matter If AI Gets It Wrong?
Because skin cancer is one of the most treatable cancers. When it’s caught early.
When it’s not? The story changes.
Melanoma in particular can spread quickly.
A spot that looks “probably fine” to ChatGPT in March can look very different by August.
After it’s had months to grow unchecked because you thought you already had your answer.
That’s not a scare tactic. That’s just how melanoma works.
The false reassurance AI provides isn’t neutral.
It actively delays care. And delayed care, with skin cancer, has real consequences.
Does This Mean AI Has No Place in Dermatology?
Some doctors are beginning to experiment with AI as a clinical tool. Something that assists a trained physician, not replaces one.
Dr. Yentzer isn’t among them. Not because he’s behind the curve – he’s published over 70 research papers and stays current on everything happening in dermatology.
But because the data doesn’t support it yet. Not in his exam room. Not with his patients.
When you come to Finger Lakes Dermatology, the person reading your skin isn’t a robot.
It’s a doctor who’s spent his career learning exactly how to do that.
What Should You Do If You Notice Something on Your Skin?
The answer is simple, even if it’s not always easy: call a dermatologist.
Not because Googling is wrong. Not because curiosity is bad.
But because the only thing that can actually tell you what’s on your skin – with accuracy, with certainty, with a plan – is a doctor who’s trained to look.
That anxious, not-knowing feeling? It doesn’t go away when AI says you’re fine.
It goes away when someone who actually knows looks you in the eye and tells you.
Have a spot you’ve been meaning to get checked?
Don’t let AI give you a false answer. Call Finger Lakes Dermatology at (607) 708-1330 or request an appointment online. We’ll take a look – and we’ll give you an actual answer.
Reference: Gandhi, I. et al. Pigmented lesion accuracy by ChatGPT in diagnosis of dermoscopic images. Journal of the American Academy of Dermatology. 2026: 94(3), 946–947.
Drs. Yentzer and Silvers at Finger Lakes Dermatology are board-certified dermatologists serving Ithaca, Cortland, and the surrounding Finger Lakes region. Here, you are a person, not a number. When you see one of our doctors, your concerns are heard. Your input is valued. And you leave feeling seen, validated, and in control of your health. Ready for that experience? Request an appointment.
This content is for informational purposes only and does not constitute medical advice. If you have a skin concern you’d like to discuss, please contact your board-certified dermatologist.