What are the FDA or legal rules regarding AI rendering of cosmetics and supplements?

Quick Answer

Under US federal law, there is no separate FDA rule for AI-rendered cosmetic or supplement photos. FDA rules govern product classification and labeling, while FTC rules require every express or implied advertising claim, including one communicated by an image, to be truthful, not misleading, and adequately substantiated. An "AI-generated" or "simulated" disclosure does not cure an image whose overall message is deceptive.

Not legal advice

This article gives general information about US federal rules, not legal advice. State laws, marketplace policies, and laws outside the US may impose additional disclosure, likeness, labeling, or advertising duties. Ask qualified counsel about the places where you sell or advertise.

Which US federal rules apply to AI images of cosmetics and supplements?

US federal law applies according to what the image communicates and where it appears, not whether AI, CGI, retouching, or a camera produced it.

Area	Federal rule for the finished image
Cosmetic labeling	FDA requires cosmetic labeling claims to be truthful and not misleading. A claim that a product treats or prevents disease, or affects the body's structure or function, can make it a drug or both a cosmetic and a drug under the FD&C Act. FDA lists acne treatment, wrinkle removal, and hair restoration among claims that have triggered warning letters for products marketed as cosmetics.
Dietary supplement labeling	A supplement may use a qualifying structure/function claim if the marketer has substantiation, uses the required disclaimer, and notifies FDA within 30 days after first marketing the product with that claim. It may not claim to diagnose, treat, cure, or prevent disease. See FDA's dietary supplement claims guide.
Advertising	The FTC evaluates an ad's overall impression, including pictures, names, captions, and omissions. Objective health or safety claims generally need competent and reliable scientific evidence before publication. See the FTC Health Products Compliance Guidance.

Which AI-rendered details create the greatest legal risk?

Generated results and before-and-after scenes. A smoother face, fuller hair, changed body, stronger immune "shield," or other visual outcome can make an efficacy claim without words. The FTC gives before-and-after imagery as an example of an implied health claim. Calling the image "simulated" cannot rescue a message that the product does not achieve or that lacks adequate evidence.
Ingredient and science cues. Fruit, botanicals, lab coats, medical symbols, and molecular graphics contribute to the ad's overall impression. Showing aloe beside an aloe-free cosmetic can imply a false composition claim; showing a sneeze blocked by a supplement can imply disease prevention.
Altered package information. The physical product still needs a compliant label. A product photo does not have to show every label panel, but a render should not change the product identity, amount, ingredient list, Supplement Facts, directions, warnings, allergen statements, or claim disclaimer and then present that package as real. FDA's cosmetic and dietary supplement guides describe the required label information.
Synthetic testimonials. An AI person used only as a visual model is not automatically a testimonial under the FTC's federal rule. Presenting a nonexistent person as a real customer, expert, or product user is different: the FTC's Consumer Reviews and Testimonials Rule covers reviews and testimonials that misrepresent that they come from someone who does not exist or lacked actual experience. An AI disclosure is not permission to invent an experience. See the FTC's rule Q&A.

How should a brand review an AI image before publishing it?

Review the complete placement, not the image in isolation: the render, caption, product name, landing-page copy, testimonials, and nearby disclaimers combine into one message. Compare the render with the physical product and the evidence supporting every visible or implied benefit. If a disclosure would contradict the main visual claim, change the visual claim.

Nightjar lets a brand group multiple photographs of its real packaging into a reusable Product, giving each Generation more evidence about the item. Built-in visual review can retry obvious product substitutions and readable text or brand-mark failures at no extra Credit cost, which reduces avoidable drift but does not establish legal compliance. A person should still compare every final image with the source product, approved claims, ingredient list, and required warnings before publication.