Products are selected by our editors, we may earn commission from links on this page.
Of 38 Korean skincare and “glass skin” videos pulled from TikTok by a team at the European Academy of Dermatology and Venereology, roughly 95 percent contained at least one potentially misleading dermatological claim. Not one was made by a dermatologist. Not one cited an outside source or reference of any kind. Lead author Dr Maria Lucia Paysano said the absence of basic safety information was the part that stood out. One video out of 38 mentioned the risk of irritation. Two mentioned sun protection. None mentioned contact dermatitis, and none suggested seeing a dermatologist. The trend promises clear, smooth, deeply hydrated skin. The study measured what the videos leave out on the way there.
This article was created with the assistance of AI and reviewed by our editorial team for accuracy and clarity.
The problem is not Korean skincare, the researchers say
The obvious conclusion from a 95 percent figure is that the products are the hazard. The researchers say the opposite. Korean skincare itself is not what they are criticizing. What they object to is the one-size-fits-all routine, handed to millions of people with different skin on the assumption that it will behave the same way on all of them. That distinction matters because it changes what a reader is supposed to do about it. Nothing in the findings argues for throwing out a cleanser. The argument is about method: whether a routine was chosen for a specific face, or copied whole from a stranger who has never seen it.
How the team built the sample and what they counted
The researchers opened a fresh TikTok account before collecting anything, a step taken to limit algorithmic bias from a personalised feed. They then searched for Korean skincare and glass skin content and gathered 38 videos, scoring each one for dermatological claims, safety information, quality and reliability. The authorship breakdown came out of that scoring. Influencers made 82 percent of the videos. Beauty brands made 11 percent. Other beauty content creators made the remaining eight percent. Dermatologists made none. The Daily Mail asked the researchers which specific videos were assessed; that detail was not in the published findings.
TikTok says 38 videos cannot describe 100 million daily posts
A TikTok spokesperson told the Daily Mail that drawing conclusions from 38 videos on a platform carrying more than 100 million posts a day is a deeply flawed methodology. The company said it proactively removes harmful health misinformation and works with fact-checkers, public health authorities and medical experts to connect users with reliable medical content. The scale gap is real and worth holding onto. Thirty-eight is a sample of what surfaced for one set of search terms, not a census of the platform. What the sample can support is narrower: among the glass skin videos a new account found, the safety gaps were close to universal rather than occasional.
A second study counted 94 patients who arrived with damaged skin
Separate research presented through the same academy looked at the other end of the pipeline. Dr Ouissal Hormi and colleagues assessed 94 patients who developed a new skin condition, or worsened an existing one, after following advice promoted on social media. Hormi said the striking part was not that the products were popular but how they were being used. Patients were reproducing complex routines with several active ingredients at once, often on the advice of people with no medical training. Among that group, 48 percent had adopted an entirely new routine and 63 percent were using more than three active ingredients at the same time.
Chemical exfoliants led the list of products involved
Chemical exfoliants containing alpha-hydroxy acids or beta-hydroxy acids, meaning acids that loosen dead surface cells, were implicated in 34 percent of the 94 cases. Over-the-counter retinoids, vitamin A derivatives that speed cell turnover, came next at 27 percent. Vitamin C serums accounted for 14 percent and niacinamide products for 11 percent. None of these are fringe items. They are the four categories a glass skin routine is most likely to stack, and three of the four work by accelerating or stripping the outer layer of skin. Used one at a time they are ordinary. The pattern in the patient data is what happens when they arrive together.
Irritant contact dermatitis was the most common reaction
The reactions were tallied by diagnosis. Irritant contact dermatitis, inflammation from direct chemical damage rather than allergy, accounted for 42 percent of cases. Acne followed at 28 percent, then rosacea or rosacea-like dermatitis at 12 percent, post-inflammatory hyperpigmentation at 11 percent and superficial chemical burns at 7 percent. Two of those entries deserve a second look. Acne and hyperpigmentation are often the reason someone starts an aggressive routine in the first place, which means a sizable share of patients ended up with a worse version of the problem they were treating. Chemical burns at 7 percent is a small number attached to a serious outcome.
Stacking actives stresses the skin barrier
Hormi’s explanation for the pattern is mechanical. Combining several active ingredients places excessive stress on the skin barrier, because some of those ingredients have overlapping irritant effects. The barrier is the outermost layer of skin, a mix of cells and lipids that holds water in and keeps irritants out. Damage it and the sequence is predictable. Water escapes faster, the skin dries out, and compounds that would normally sit on the surface reach deeper tissue. Products that felt tolerable individually start stinging. That is the point at which many people in the patient group concluded their routine was not strong enough and added something else.
What the first study found missing from the videos
Paysano’s team recorded specific omissions rather than general vagueness. Eighty-four percent of the 38 videos failed to account for different skin types. Forty-five percent promoted excessive combinations of active ingredients without appropriate safety guidance. Thirty-two percent promised instant results. Those three numbers map directly onto the injuries in the second study: the wrong routine for a given face, too many actives at once, and an expectation of speed that encourages escalation. Paysano described a clear gap between how popular the content is and how good the information in it is. Following these routines without individualized guidance, she said, may contribute to irritation, over-exfoliation, acne flares or barrier disruption.
Both authors say dermatologists should join the conversation
Hormi’s summary was one sentence: viral does not necessarily mean safe, validated or suitable for everyone. Paysano went further, arguing that dermatologists have a role beyond warning people off. Rather than dismissing social media trends, she said, the profession has an opportunity to take part in them and make evidence-based skincare information accurate, accessible and engaging. It is an unusual position for a specialist body to take about a platform its own research just criticized. The practical advice is narrower. Be cautious of content promising instant or universal results, and if you have an existing skin condition, develop irritation, or are unsure whether a routine suits you, ask a dermatologist.