Fred: A Supplement Assistant That Won’t Turn Your Product Into a Drug

A shopper asks your supplement brand’s website whether a product will cure their anxiety and lower their blood pressure. A general-purpose chatbot says yes to both. In two sentences it turned a dietary supplement into an unapproved drug, because a claim to treat or cure a disease is a drug claim under federal law. The brand wanted a tool to help shoppers and support orders. It got one crossing the exact line that defines the category.

The threat and standard articles in this series walk through how one verb does that. This one is the answer: what a compliant AI assistant for supplements looks like, and why the design, not a disclaimer, is what keeps a website from making disease claims, handing out doses, or clearing interactions.

The Real Choice Is Governed or Ungoverned

Brands ask whether to put AI on the site at all. Shoppers already decided. They want answers about products before they buy. The real question is whether the assistant is governed.

An ungoverned chatbot promises cures, invents benefits, and recommends doses. A governed one describes products within the rules, keeps claims substantiated, and routes medical questions to a professional. Same convenience, completely different exposure when a disease claim shows up in a transcript.

How Fred Holds the Line

Fred starts from the opposite default of a general chatbot. A generic bot answers anything unless told not to. Fred answers only what it is cleared to, and the boundary lives in the system rather than a prompt that drifts when a shopper names a condition.

In practice, Fred works from your own approved content. It describes products, ingredients, and orders. It does not say a product treats, cures, or prevents a disease, because that turns it into a drug under federal law; it keeps to the limited structure-function statements supplements may make. It does not invent benefits or overstate evidence, because unsupported health claims fall under the FTC’s deceptive-practices authority. And it does not give a dose or clear a drug interaction. Fred runs more than 50 industry guardrail packs, and the supplement pack is built around the structure-function versus disease-claim line, substantiation, and the boundary on medical advice.

The design proves itself under pressure. A shopper will ask "will this help my arthritis pain?" several ways, and a prompt-instructed bot eventually says yes, because a named condition slips past the instruction. Fred does not rely on catching the phrasing. It decides what may be said before the reply forms, so a disease claim never reaches a shopper. "Will not" is a suggestion. "Cannot" is an architecture.

Fred vs. a Generic AI Chatbot

That single screen is the argument. A general chatbot is encouraging until encouraging becomes a disease claim. Fred is helpful across everything that supports a sale and structurally incapable of turning your product into an unapproved drug.

What Your Brand Actually Gets

Set the rules aside and look at the store. Fred handles the after-hours questions that usually go unanswered, what’s in this, how is it made, what’s your return policy, where’s my order, and answers them instantly. It supports the purchase and captures the shopper, and routes disease-claim questions and medical questions to a professional.

The captured-sale math is the real win. The shopper who gets a fast, accurate, compliant answer at 9 p.m. is the one who buys, instead of the one who left. Fred turns the after-hours window into orders without ever putting a disease claim or a dose on the record.

So the question is not whether supplement brands will use AI. They will. It is whether yours stays inside the claim rules, every time, no matter how the question is asked.

Frequently asked questions