Cosmetics have been the most notified category on the EU's Safety Gate system for three years running, accounting for 36% of all alerts in 2025, at close to 100 alerts a week across categories. That figure frames the current debate over beauty dupes better than any argument about morality or price, because detection of non-compliance is now automated and rising.

The dupe debate found fresh fuel recently when Charlotte Tilbury told the BBC: "when you dupe, you dupe the consumer." She added: "I have looked at those dupes, and they do not [perform at the quality that Charlotte Tilbury products do]. If I could do them for those prices, I would."

Lee Bryan, founder and CEO of Arcus Compliance, counters that dupes, done properly, are entirely legal and can be just as compliant as the originals. The real dividing line, he argues, is not price or positioning but the state of the regulatory file behind the product.

"Dupes are often perceived as copying someone else's work, fairly or not, so these brands tend to start with less benefit of the doubt than others on the shelf. That is exactly why compliance has to be watertight," Bryan told Personal Care Insights.

What does 'dupe' actually mean to a regulator?

Bryan draws a sharp distinction the industry often blurs. A counterfeit pretends to be the original and infringes intellectual property. A dupe competes openly on similarity and, done properly, is legal. The regulator, he says, does not need to settle the philosophy. "They ask one question, which is 'Show me the file.'"

The structural risk is in the claim, not the intent. The moment a brand positions its product against a prestige original, it imports that product's performance promise without importing any of the evidence. Under the Common Criteria, evidential support is required for what a claim implies, not simply what it states — and the fairness criterion explicitly prohibits creating confusion with a competitor's product.

The problem compounds because most dupe brands did not develop their own formulation. They lean on a manufacturer's dossier that may have been assembled for a different claim set, a different concentration, or a different market. Bryan's test is blunt: if a regulator asks for the study behind your headline claim tomorrow morning, do you hold it, or are you emailing your supplier and hoping?

Where do identical products come from?

Bryan describes a pattern he sees constantly in the market:

• A founder borrows money from family and finds a white-label manufacturer
• They pick a base formulation out of a catalog and put their brand on it
• Six months later, another founder does the same, at the same manufacturer, off the same base
• Neither copied the other; neither owns the formulation — the contract manufacturer does

The result is two chemically identical products on sale under different names, at different prices, with different brand stories. If the product is identical, Bryan argues, the product cannot be the differentiator. What separates the survivors is diligence on the manufacturer, a commissioned safety assessment, and ownership of the file.

That distinction is commercial, not just ethical. When enforcement arrives, one of two identical brands turns an inspection around within a week because its file is complete and current. The other goes quiet, panics, withdraws stock and ends up in a recall. Same molecules, same factory, potentially the same batch — completely different outcome.

Has enforcement shifted under dupes?

Yes, and by automation. The EU's eSurveillance crawler inspected more than 1.6 million URLs last year, and a second proactive crawler has now launched that actively searches for problems rather than matching listings against existing alerts. Bryan cautions against reading rising notifications as a suddenly more dangerous market: detection is rising, so notifications rise with it.

The UK picture adds another layer. UK and EU annexes have diverged since Brexit, so a brand selling into both markets on a single dossier is exposed in at least one of them without knowing it.

Cost-down reformulation compounds the documentation gap. Every reformulation is a new safety assessment, yet in practice brands change the fragrance house, swap a raw material supplier, shave a percentage, and never reopen the report. The file ends up describing a product that no longer exists.

Do influencer captions carry regulatory risk?

The legal exposure, Bryan says, has quietly moved off the label and into the caption, while cosmetics regulation was built around the pack. Two traps catch founders:

• An implied comparison is still a comparison. Saying a product works just like a named prestige original is a performance claim needing evidence, a comparative advertising claim in its own right, and a direct run at the fairness criterion
• Ownership of the claim. If a brand briefed it, gifted it, paid for it or approved it, it is that brand's claim. Brands file influencer output under earned media; regulators file it under advertising the brand controls

The enforcement backdrop has hardened too. Since April 2025, the UK's Competition and Markets Authority can itself determine consumer law breaches and impose penalties of up to 10 percent of global turnover without going to court. Most beauty founders, Bryan says, have not absorbed what that means for the content they commission.

Spoken claims are no longer ephemeral. A 30-second video is machine transcribable, and the same automation driving Safety Gate volumes reads captions perfectly well. A claim never written down is now indexed, searchable and permanent — and if a consumer positioned on equivalence experiences anything less, the brand has engineered its own complaint pipeline.

What should dupe brands do now?

Bryan rejects the premise that cheap means unsafe. "Undocumented means unsafe," he says, noting some of the best-evidenced products on the market are inexpensive, while some of the worst-documented carried a three-figure price tag. His operating standard is that compliance you cannot evidence within a week is not compliance; it is a belief. His recommendations:

• Hold a live claim register, mapping every claim in every market to its supporting evidence, with social and influencer content inside the register, not outside it
• Version control the safety file so a formulation change automatically triggers a dated, recorded reassessment
• Screen ingredients against each destination market before launch, not after a query
• Run post-market surveillance that genuinely exists: an adverse event log and a rehearsed recall procedure
• Bring the Responsible Person in before the campaign goes live, not after the regulator calls

Safety documentation, Bryan stresses, is where dupe brands most often fall short — usually on evidence rather than intent.

The next two years, in his assessment, will not reward the brands that argue loudest that they are legitimate. They will reward the ones that can prove it on a Tuesday morning, at short notice, to a system that found them automatically.