Accelerating into fraud
So far, AI is giving us more of the same
Last week, the New York Times breathlessly reported that a startup telehealth provider was using AI to run most of its business. In fact, the story explained, MEDVi has only two employees — its founder and his brother. What’s more, this company is on track for $1.8 billion in sales in 2026. They’re making that money as a dropshipping service for GLP-1s, using a white-labeled telehealth provider called OpenLoop.
The New York Times wrote about this company almost as a scaremongering tactic, making the company into a symbol of what we can expect when AI takes over everything.
A $1.8 billion company with just two employees? In the age of A.I., it’s increasingly possible…Mr. Gallagher does not anticipate hiring more people. He said he just didn’t see how it would help Medvi [sic], though he misses the camaraderie of colleagues.
“At this point, I kind of want to hire people because I’m lonely,” he said.
The problem is that this company is on track to do so much revenue probably because it’s committing fraud.
RICO for healthcare
Brendan Keeler has been tracking this company since January in his newsletter Health API Guy, writing then that MEDVi is named in a RICO (Racketeer Influenced and Corrupt Organizations) Act lawsuit alleging that it and several other companies were shell companies, using OpenLoop’s white-labeled service to sell oral tirzepatide that was no better than a placebo. (Because, according to the suit, tirzepatide is not a molecule that can be processed orally by the human body.)
It gets even worse. In a piece published this week, Brendan wrote:
When we first mentioned MEDVi back in January, it appeared as one of many nearly identical storefronts in the OpenLoop litigation, just a name among a list of brands sharing templates, infrastructure, and a common backend. It looked interchangeable. The recent filings suggest otherwise. Across multiple cases, MEDVi shows up not just as a participant, but as a central node connecting traffic generation, conversion, and downstream clinical and fulfillment processes.
MEDVi is also allegedly violating the Federal Trade Commission’s requirement on clearly disclosing paid advertising by actor portrayals — or in MEDVi’s case, AI portrayals of doctors on social media. The company has been sued for its AI bots and advertising affiliates (the reporting on MEDVi uses the word “affiliate” a lot, and I suspect that it’s because the reporters can’t verify if a given action is being taken by a human or an AI agent) violating spam laws, as reported by Bloomberg.
This is not a telehealth company that is providing healthcare, it’s a cash grab while compounded GLP-1s are still mostly unregulated. The humans who interact with it are funneled through trick mirror AI from the ad they see to the website they visit to the customer service agent they use. (You would think the New York Times could do a quick google search before publishing.)
Related articles
We’re moving fast and breaking things, but it’s to obscure shady business practices, sell more people more compounded weight loss drugs, and pocket the cash.
Accelerating the worst parts of the healthcare system
I am once again confronted with a recurring question: is AI creating anything net-new, or is it exacerbating existing issues? I’ve been writing some version of this question for about 6 months now. In healthcare at least, it’s mostly exacerbating existing issues.
This isn’t me saying that AI can’t be transformative or that it should be dismissed as merely a tool to commit fraud. (Although the recent JAMA study showing that AI scribes aren’t saving meaningful time so far is jarring — as is the conclusion that this means providers using the scribes can see one additional patient every 2 weeks.)
It is me saying that AI dropped into our current healthcare system — broken in all the ways that are familiar, from misaligned incentives to loss of trust in physicians — is causing humans to react in predictable but harmful ways.
AI is accelerating, but it’s accelerating our broken system.