Skip to main content
AI-Augmented Audits 9. Juli 2026

When a Cosmetic Becomes a Drug: How Marketing Claims Trigger FDA Oversight — and What Manufacturers Must Audit

One marketing claim can reclassify your cosmetic as an unapproved drug. Learn what triggers FDA's intended use doctrine and how to audit your full claim inventory.

SS
Sam Sammane
Founder & CEO, Aurora TIC | Founder, Qalitex Group

A moisturizer becomes a drug the moment its label says it “rebuilds collagen.” That’s not a metaphor — it’s a legal reality under the Federal Food, Drug, and Cosmetic Act, and it carries consequences most cosmetic formulators and brand marketers never see coming.

This classification boundary is one of the most commercially dangerous in regulated product development. It determines which regulatory framework applies, which labeling requirements kick in, and which FDA enforcement division is watching your product. Getting it wrong doesn’t just mean a Warning Letter. It means your product is an unapproved new drug on the market — a different category of problem entirely, with potential for seizure and injunction proceedings, not just voluntary correction.

The Statutory Line: Cosmetic vs. Drug Under 21 U.S.C. § 321

The FD&C Act draws the boundary using a deceptively simple concept: intended use. A cosmetic (21 U.S.C. § 321(i)) covers articles intended to be applied to the human body for cleansing, beautifying, promoting attractiveness, or altering appearance. A drug (21 U.S.C. § 321(g)(1)) includes any article intended to affect the structure or any function of the human body.

The critical phrase is “intended to.” And FDA determines intended use primarily through what labeling, advertising, and marketing materials say — not what the formulation actually does at a biochemical level.

This is why two nearly identical moisturizers can have entirely different regulatory status. One says it “visibly reduces the appearance of fine lines.” The other says it “stimulates cell turnover to rebuild collagen.” The first is a cosmetic. The second is a drug, because it claims a physiological mechanism of action.

FDA codified its intended use standard at 21 CFR § 201.128, and the scope there is broader than most people expect. Intended use may be established through labeling claims, advertising matter, or oral and written statements by anyone in the distribution chain. That last clause matters. A sales representative’s email or a brand ambassador’s social post can contribute to FDA’s determination that a product makes drug claims — even if the physical label is technically clean.

The Claims That Cross the Line

There’s no single FDA-published list of forbidden cosmetic words. Instead, FDA evaluates the totality of the evidence, which means you need to understand the underlying principle to audit effectively rather than just memorize a phrase list.

Claims that typically remain within cosmetic territory:

  • “Visibly reduces the appearance of wrinkles”
  • “Leaves skin looking firmer”
  • “Helps skin retain moisture”
  • “Brightens skin tone”

Claims that typically trigger drug classification:

  • “Rebuilds collagen” or “stimulates collagen synthesis”
  • “Increases cellular turnover rate”
  • “Clinically proven to reduce wrinkle depth by 43% in 8 weeks”
  • “Relaxes facial muscles” (this alone invites a Botox-comparison analysis)
  • “Controls sebum production” — particularly when framed as a mechanism rather than an appearance effect
  • “Accelerates wound healing” or “repairs the skin barrier”

The OTC drug monograph system adds a third tier. Products like sunscreens, anti-dandruff shampoos, and fluoride toothpastes are simultaneously cosmetics and drugs — they must comply with OTC monograph requirements, including active ingredient disclosures, Drug Facts labeling panels, and specific usage directions. There are currently more than 25 established OTC drug monograph categories, each with its own permitted active ingredient concentrations and labeling requirements. Missing the Drug Facts panel on an SPF moisturizer is a routine audit gap, even among sophisticated brands.

The Modernization of Cosmetics Regulation Act (MoCRA, P.L. 117-328, signed December 29, 2022) tightened the cosmetic regulatory framework significantly — facility registration, adverse event reporting for serious events, and safety substantiation requirements all apply under the new law. But MoCRA didn’t change the cosmetic/drug boundary itself. Products that make drug claims still route to CDER enforcement, not to OCC cosmetics oversight.

What a Boundary Audit Actually Looks Like

If you’re managing a personal care or cosmetic-adjacent health product portfolio, a defensible claim audit covers at least five distinct domains. Each one has its own failure mode.

1. Label Copy Review Against the Intended Use Doctrine Every claim on the primary display panel, inner carton, and secondary labeling gets mapped against the cosmetic/drug distinction. This includes directional text, benefit statements, and embedded scientific language. “Dermatologist-tested” isn’t inherently a problem — but “dermatologist-recommended for treating dry skin conditions” edges into disease claim territory, which is a separate problem from the drug claim issue.

2. Digital Marketing and Website Content FDA’s position since the early 2000s is that website claims about a product are labeling. That includes Amazon product pages, Walmart.com listings, and brand DTC sites. A product whose physical label is clean but whose Amazon A+ content page says it “restores barrier function compromised by eczema” has a significant compliance exposure. A mid-size brand might have 200+ pages of product-related web content distributed across channels — and manual review of that volume is genuinely difficult to sustain.

3. Third-Party Distributor and Retailer Materials Under the intended use standard, retailer-generated claims can contribute to FDA’s classification analysis. This catches many brands off guard. A domestic distributor who rewrites product descriptions for their own catalog, or a European retailer whose localized website adds mechanism language not present in the original label, can create U.S. regulatory exposure for the brand owner.

4. Clinical Study References Citing clinical data triggers immediate additional scrutiny. A claim like “shown to reduce wrinkle depth by 37% over 12 weeks” sounds scientific, but it raises immediate questions: Was the study IRB-approved? Did the endpoint measure cosmetic appearance or physiological change? If the study measured transepidermal water loss, collagen density by biopsy, or any biomarker of structural tissue change, the claim is almost certainly a drug claim — regardless of how the brand frames it.

5. Combination Product Identification Some products genuinely function as both cosmetics and drugs: sunscreens with moisturizing claims, anti-dandruff conditioners, fluoride-containing mouthwashes. These must satisfy both regulatory frameworks simultaneously — Drug Facts labeling and OTC monograph active ingredient requirements on the drug side, plus 21 CFR Part 701 ingredient disclosure on the cosmetic side. Missing one side of that equation is one of the most common gaps we see in combination product portfolios.

Where AI Changes the Calculus for Claim Auditing

The volume problem in marketing claim review is real. Consider a brand with 150 SKUs, active across 8 retail platforms, with 2–3 variant listings per product. That’s potentially 3,000+ pages of product content that theoretically need claim-level review. No human audit team touches that comprehensively at sustainable cost or speed — not even close.

This is where AI-augmented audit tools are showing genuine operational value. Not as a replacement for regulatory expertise, but as a first-pass filter that surfaces high-risk content for expert review.

The approach we use in our consulting workflows at Aurora TIC applies natural language processing to classify marketing claims by risk tier. Inputs can include scraped web content, label images processed through OCR, or direct copy feeds from content management systems. The output is a structured flags inventory — every instance where a claim uses mechanism language, disease terminology, clinical study references, or structure/function framing that typically triggers drug classification.

The AI system doesn’t make the regulatory call — that still requires a human expert with command of the applicable OTC monograph categories, the specific product category’s enforcement history, and FDA’s current warning letter posture in the relevant space. But it compresses what would otherwise be a 40-hour manual audit to something closer to 4 hours of expert review time. And it catches the things that manual review misses: the archived blog post on a forgotten subdomain, the discontinued product whose Amazon listing still gets indexed by FDA’s market surveillance tools.

In structured audits running this workflow, roughly 1 in 6 cosmetic-adjacent products has at least one instance of language that would draw FDA scrutiny under the intended use doctrine. That rate is high enough to make systematic scanning — rather than periodic spot-checking — the defensible standard of care for any serious compliance program.

The Case for Proactive Regulatory Compliance Consulting

FDA’s enforcement posture for drug claims in cosmetic products has historically been reactive — triggered by adverse event reports, competitor tip-offs, or media coverage. But that’s changing. FDA’s expanded cosmetic authorities under MoCRA, combined with the agency’s increasing use of automated market surveillance tools, suggest that detection rates will increase over the next several years. FDA has made no secret of the fact that it considers the “cosmeceutical” marketing category to be a regulatory fiction.

The brands at greatest risk right now are those in the clinically-active skincare and “active ingredient” personal care segments — products positioned as delivering real physiological outcomes (cell renewal, collagen regeneration, barrier restoration) through marketing language, while relying on cosmetic-tier regulatory treatment. This positioning is commercially attractive. It’s also the exact profile that drives Warning Letters from CDER’s Office of Compliance.

A proactive approach means auditing your full claim inventory now, before an adverse event report puts a specific SKU on FDA’s radar. That means covering not just current labeling but archived web content, influencer brief language, and retailer marketing guidelines that your team may not have reviewed in years.

If your product portfolio sits anywhere near the cosmetic/drug boundary — and in the modern “active ingredient” skincare era, many do without realizing it — a structured regulatory compliance consulting engagement is worth serious consideration. The cost of a thorough audit is a small fraction of a consent decree. And it’s a far easier conversation to have with your leadership team before the Warning Letter than after.


Written by Sam Sammane, Founder & CEO, Aurora TIC | Founder, Qalitex Group. Learn more about our team

Reserve early access to our AI audit tools Contact us

Benötigen Sie Hilfe bei der Auswahl des richtigen Labors?

Aurora TIC verbindet Hersteller und Marken mit akkreditierten Prüflaboratorien — schnell, kostenlos und auf Ihr Produkt zugeschnitten.

Kostenloses Angebot anfordern