1,046 Fake AI Doctors Are Flooding YouTube With Medical Lies

The scale nobody was watching for

"Dr. Richard Ben." "Dr. Laura Mendoza." "Dr. John Meyers." Names built to sound credible, attached to faces built to sound the same way. None of them belong to a licensed physician. They're synthetic personas, and Maldita.es found at least 1,046 channels like them, spread across more than thirty languages.

The oldest account traces back two decades, but that's the exception. Over three-quarters of these channels appeared after January 2025 — right as avatar generation and voice cloning got cheap enough for anyone to run a content farm out of a laptop. This isn't a fringe problem anymore. It's an industry, and it runs on the same AI video generation that makes deepfakes hard to spot at a glance.

The claims are the dangerous part

A thumbnail from "Dr. Richard Ben" claims cancer can vanish within 24 hours. Another, from "Dr. Laura Mendoza," promises Alzheimer's disappears if you eat the right fruit. A third pushes the same line for diabetes. None of it holds up. There is no fruit, no supplement, no 24-hour fix for any of these diseases — and pretending otherwise isn't harmless content, it's a health risk aimed at people who are scared and looking for hope.

Maldita's team went further and tracked 20 channels making direct cancer-cure claims. They reported the videos to YouTube. Seventy-two hours later, every single one was still live. That gap between "reported" and "removed" is where the damage happens — a family member searching for answers at 2 a.m. doesn't wait for moderation to catch up.

How the fake credibility gets built

The formula repeats itself channel after channel, language after language: scrubs, a stethoscope, a generic exam-room background, and a bio claiming "25 years of experience" or a promise to reveal what "nobody else tells you." A few channels tack on a disclaimer about not replacing professional medical advice. It's usually buried well below the thumbnail that already made the false promise — and the thumbnail is what gets the click.

Why it keeps working

Money is the throughline. Maldita found that 45% of these channels monetize directly through YouTube's ad revenue share. Others run paid memberships priced up to €25 a month, or link out to donation pages and supplement sales. One French-language channel sells a wellness guide for about €8. None of this requires the "doctor" behind it to exist — the avatar just needs to be convincing enough to hold attention for the length of a pre-roll ad.

It also spreads sideways. A related case reported by the BBC involved a UK hospital trust issuing a public alert after AI-generated clips falsely showed its own clinicians endorsing weight-loss patches on Facebook and TikTok. A hospital official put it plainly: AI-generated images can now be very good — good enough that a real medical staff member's likeness got hijacked without anyone noticing until patients started asking questions. That's the same category of fraud our AI scam and deepfake detector is built to catch before it reaches someone who trusts it.

What actually protects you

Frequently asked questions