We first covered the stock photo doctor problem in early 2026. Since then, it's gotten both better and worse. Better: several platforms we called out removed fake team photos. Worse: the fakes got harder to spot, because AI-generated portraits have reached a quality level where casual visual inspection doesn't catch them anymore.
This updated guide covers three categories of fake medical team imagery — stock photos, AI-generated portraits, and "borrowed" photos from real doctors who don't work at the platform — and the current best tools for catching each one.
Category 1: Stock photos (still the most common)
Stock photo doctors are the original sin of telehealth marketing. The same smiling woman in scrubs appears on six different platforms because all six licensed the same Getty or Shutterstock image. She's not a doctor. She's a model. And the platform using her photo is betting you won't check.
How to catch them
Google Lens. Right-click the photo on the platform's website. Select "Search image with Google Lens" (or upload to images.google.com). If the same face appears on multiple unrelated sites — a dental practice in Tampa, a dermatology clinic in Seoul, a GLP-1 platform in Delaware — it's a stock photo. This takes approximately fifteen seconds per image.
TinEye. Upload the image to tineye.com. TinEye specializes in finding exact and near-exact matches across the web. It's particularly good at catching images that have been cropped, color-shifted, or horizontally flipped — common modifications platforms use to make stock photos harder to trace.
Stock photo detection is still the easiest audit. It takes under a minute per image, and false positive rates are near zero — if the same portrait appears on a stock photography marketplace, it's a stock photo.
Category 2: AI-generated portraits (the growing threat)
In 2024, AI-generated faces were detectable by artifacts: mismatched earrings, teeth that blurred into each other, backgrounds that bent around the head. In 2026, the best generators produce portraits indistinguishable from real photographs at web resolution.
The tell is no longer in the pixels — it's in the metadata and the ecosystem around the image.
How to catch them
Reverse-image search returns nothing. This is the primary signal. A real person's professional headshot typically appears in at least one other context — a LinkedIn profile, a hospital directory, a medical school graduation page. An AI-generated portrait exists in exactly one place: the platform that created it. If Google Lens and TinEye return zero results for a professional headshot, that's suspicious. Real doctors leave digital footprints.
Cross-reference the claimed name. If the photo accompanies a name (e.g., "Dr. Sarah Mitchell, MD"), search that name on the NPI Registry. If no NPI exists for that name and specialty, or if the NPI photo (where available through state board profiles) doesn't match the website photo, you've found a fabrication.
Check for C2PA metadata. The Coalition for Content Provenance and Authenticity (C2PA) standard embeds origin metadata in images. Some AI generators now include C2PA credentials. Tools like Content Credentials Verify can detect this. It's not universal yet, but when present, it's definitive.
Category 3: Borrowed photos (real doctors, fake affiliations)
This is the most insidious category. Some platforms use photos of real, licensed physicians who have no relationship with the platform. The doctor's photo is pulled from their hospital bio page or LinkedIn profile and placed on the telehealth platform's site without consent.
We've documented cases where physicians contacted us after discovering their photos on platforms they'd never heard of. In at least two instances, the physicians had pending malpractice concerns that the platform was unknowingly inheriting by association.
How to catch them
Reverse-image search, then compare contexts. If the same headshot appears on a hospital system's provider directory AND the telehealth platform, check whether the platform's claims match. Does the hospital page list the physician as an employee? Does the telehealth platform claim them as their medical director? If the physician works full-time at a health system, they're probably not also running a GLP-1 startup's clinical operations.
Verify directly. If a platform names a specific physician as their medical director, you can verify through the physician's own channels — their practice website, their hospital bio, or a direct call to their listed office. If they don't know they're listed, that tells you everything.
What legitimate platforms do differently
Platforms that pass our audit consistently share three traits in their team presentation:
Named clinicians with NPI numbers. Not stock art. Not AI portraits. Real names attached to verifiable national provider identifiers. The NPI is the universal key that unlocks state license verification, specialty confirmation, and disciplinary history.
Inconsistent photography. This sounds counterintuitive, but real medical team pages look slightly messy. Different headshot backgrounds, different lighting, maybe one person's photo is clearly from a phone camera while another is professionally shot. Uniform perfection is a design choice; clinical authenticity is inherently imperfect.
Active clinician profiles elsewhere. A real medical director has a LinkedIn profile, a state medical board listing, perhaps a Doximity profile or published papers. They exist in the world beyond the platform's marketing page.
The 60-second audit
You don't need to do all of the above for every platform. Here's the minimum viable audit, which takes roughly one minute:
Step 1 (15 seconds): Go to the platform's "About" or "Our Team" page. Are individual clinicians named?
Step 2 (15 seconds): Pick one name. Search it on the NPI Registry. Does the result match the claimed specialty and location?
Step 3 (15 seconds): Right-click one team photo. Run Google Lens. Does it appear anywhere else?
Step 4 (15 seconds): Check the photo count. Are there more stock-looking headshots than named clinicians? Ratio matters.
If any step fails, dig deeper before you hand over your medical history and credit card. The platforms that invest in real clinical teams don't hide them. The platforms that hide their teams are hiding them for a reason.