AI telehealth ad scandal raises ethical alarm
A report revealed Medvi, a telehealth firm, used deepfake AI‑generated doctor profiles — over 800 fake clinician ads — raising fresh ethical concerns about authenticity and trust in digital health platforms. The episode highlights risks when patients can’t verify who is providing virtual care, a problem that could spill into maternal telehealth services like remote triage or lactation support. Regulators and platforms may need clearer guardrails to prevent deceptive practices in health advertising. (x.com)
Medvi grew fast by selling online weight-loss care, and Forbes reported on April 2 that the company used Midjourney and Runway to generate ad creative while running with a headcount of two. A separate report then found more than 800 ads that showed fake clinician identities, turning a telehealth marketing shortcut into a trust problem. (forbes.com) (404media.co) Telehealth works on a simple bargain: the patient cannot see the clinic, so the clinic has to be unusually clear about who is treating them. If the doctor photo, name, or credentials in an ad are synthetic, the first layer of that bargain is already broken before a visit begins. (hhs.gov) (ftc.gov) This landed in a market already moving at breakneck speed. Medvi launched in September 2024, booked $401 million in 2025 revenue on 250,000 customers, and is projected by Forbes to reach $1.8 billion in 2026 sales. (forbes.com) A lot of that market is built around glucagon-like peptide-1 drugs, the class that includes semaglutide and tirzepatide. Fierce Healthcare reported in 2025 that a STAT investigation had already identified more than 100 telehealth sites marketing these drugs, often relying on a small number of prescribing networks behind the scenes. (fiercehealthcare.com) That scale is why fake clinician ads are different from a normal internet scam. When one company can reach hundreds of thousands of patients with paid creative, a made-up doctor profile can function like a counterfeit white coat at industrial volume. (forbes.com) (ftc.gov) United States regulators were already moving toward this fight before the Medvi story surfaced. The Food and Drug Administration said on March 3, 2026 that it sent 30 warning letters to telehealth companies over false and misleading claims for compounded glucagon-like peptide-1 products. (fda.gov) The Federal Trade Commission has also been building an impersonation casebook around artificial intelligence. In February 2024 it proposed stronger protections against impersonation of individuals and said artificial-intelligence-generated deepfakes could “turbocharge” fraud at wider scale. (ftc.gov) This is not only about obesity drugs. Maternal telehealth uses the same remote trust model for triage, postpartum check-ins, and lactation support, so a patient who cannot easily verify the human on the screen is exposed to the same identity problem in a much more vulnerable moment. (hhs.gov 1) (hhs.gov 2) The practical fix is boring but concrete: real license numbers, verifiable state registrations, clear disclosure when images are synthetic, and ad platforms that treat fake clinician identities the way banks treat fake passports. Without that, telehealth keeps asking patients to trust a face that may have been rendered by software five minutes before the ad went live. (ftc.gov 1) (ftc.gov 2)
