The New York Times called Medvi a $1.8 billion company built by one person in two months using AI. The reality is messier. Medvi faces a class-action lawsuit for violating California's anti-spam laws. Affiliate marketers allegedly used spoofed domains and deceptive subject lines to evade filters. Rob Freund, an attorney tracking e-commerce compliance, flagged the lawsuit publicly. The Times verified $60-70 million in cleared revenue but didn't dig into what's actually being sold.
Investigations by Futurism and YouTuber Voidzilla paint a different picture than the vibe-coded success story. One source described Medvi as "a fraud-layer on top of also-scammy-but-possibly-less-illegal platforms." The company relies on OpenLoop and a network of contractors, not a solo founder wielding AI tools. That $1.8 billion valuation, roughly a 30x revenue multiple, ignores these dependencies.
Marcus makes a blunt point: if you're looking for an AI poster child, this isn't it. The story shows how AI agents can help scale operations that might not survive traditional scrutiny. The Times piece treated the revenue figures as validation. His question: why assume that's the only thing the company is being honest about?