Are AI skin scanners worth it?

Walk into a growing number of UK beauty clinics and you’ll be offered something that you rarely saw a decade ago: a dedicated skin imaging scan. Devices like the VISIA and OBSERV have moved from aesthetic doctors’ clinics into mainstream beauty destinations, mainly due to the demand for data-led, personalised offerings across all things beauty and wellness.

Some clinics fold the scan into their consultation fee, while others charge separately, sometimes redeemable against a treatment, sometimes not. The question is whether the data genuinely changes what practitioners recommend, or whether a good clinician’s trained eye remains the most valuable diagnostic tool in the room?

Skin analysis devices explained

The most widely used systems in UK clinics are the VISIA, and the OBSERV 520x. Both capture standardised, high-resolution photographs under multiple lighting conditions to reveal concerns both on and beneath the skin’s surface.

VISIA measures wrinkles, UV damage, pores, pigmentation, redness, porphyrins linked to acne bacteria and texture, comparing results against a reference database of people of similar age and skin profile to generate percentage scores. OBSERV operates across similar modes but without automated scoring.

These systems use machine learning, a branch of AI in which software is trained on large datasets of skin images, to identify and categorise patterns, comparing your skin against thousands of reference images to generate consistent scores. What they are not doing however, is diagnosing conditions or making clinical decisions — that remains the job of whoever is reading the results.

What they can detect that the naked eye can’t

The genuine clinical advantage of these systems is their ability to reveal what lies beneath the surface. UV damage that hasn’t yet appeared as visible pigmentation, subsurface redness, the full extent of bacterial congestion in the pores, all of these can be significantly more apparent under UV or polarised light than under ordinary clinic lighting.

They rarely move the needle in terms of what treatments Aesthetic Doctor and The Aesthetic Doctor Clinic Founder, Dr Ahmed El Muntasar, would recommend, but, “telling someone that they have accumulated UV damage can feel quite abstract. Showing them a UV image of that damage makes the conversation much more tangible,” and that’s where he finds these scanners to be most helpful. “They fill the gap between knowing you should wear SPF daily, and actually doing it,” he adds.

What they can’t tell you

“Skin is a reflection of overall health. AI cannot fully appreciate symptoms such as pain or itching, hormonal fluctuations, lifestyle factors, psychological impact or the wider medical context that often explains why someone’s skin has changed. That’s where clinical reasoning remains essential,” believes Founder and Advanced Clinical Practitioner at Tan&Co, Nurse Teresa Tan.

A scan simply captures a single snapshot in time, taken under controlled conditions that don’t account for dehydration, recent exercise, temperature shifts, products on the skin, hormonal fluctuations, medications, stress, sleep quality, nutritional deficiencies and previous treatment history.

For example, a high redness score doesn’t tell you whether you’re looking at rosacea, irritation, a reaction to a new cleanser or a natural flush. “Technology can identify patterns; clinicians determine whether those patterns are clinically significant and what, if anything, should be done about them,” says Tan.

Skin specialist Corina Mihalache, who has evaluated several systems and chosen not to introduce one, believes these devices also miss other key nuances that impact what treatments and skincare is best for you.

“I’m assessing tissue movement, subtle laxity, micro inflammation, and tactile texture, things a camera lens simply can’t do. A scanner can’t tell me how a patient’s skin feels after cleansing, how sensitive it is to active ingredients, how stress affects their breakouts, or how their barrier reacts to environmental changes,” she explains. And her preference is clear, “I prefer my patients’ care to be driven by expert clinical judgement, rather than an algorithm’s standard baseline.”

Does it change the treatment plan?

In El Muntasar’s experience, a scan is more likely to refine a treatment plan than fundamentally change it, but that refinement can matter. “A patient may come in primarily concerned about enlarged pores and uneven texture, and assume they need an aggressive resurfacing treatment. Imaging may reveal that much of the unevenness is actually related to underlying pigmentation and redness. In that situation, I may recommend starting with pigment control, daily SPF and a treatment targeting vascular or pigment changes, before moving on to more aggressive resurfacing,” he shares.

That’s no small thing, but when clinics use branded devices that are designed to only recommend one specific skincare brand — meaning automated recommendations are limited to products within that range — there’s a concern regarding objectivity in these instances that Tan believes we can’t ignore.

“I prefer to retain complete independence and select the most appropriate treatments and products for each patient, rather than allowing an automated system or brand affiliation to influence the plan,” agrees Mihalache.

The skin tone problem

One concern that rarely features in the marketing around these devices is how consistently they perform across different skin tones. Machine learning systems are only as reliable as the data on which they were built, and if darker skin tones are underrepresented in the training images, the system may struggle to accurately distinguish normal variation in skin colour, from pigmentation to inflammation.

“We already know from wider dermatology research that some AI models perform less accurately on darker skin. I would therefore want to know whether a system has been properly validated across a genuinely diverse range of skin tones, before placing too much weight on its scores,” says El Muntasar. It is a question worth asking before booking in for a scan or when being offered one, even if it’s complimentary.

“I would introduce one if I felt it measurably improved treatment decisions across all the skin tones we treat, not simply because it appeared impressive during a consultation,” says Mihalache.

Is it worth paying for?

The answer depends entirely on what’s included in the fee and who is interpreting the results. “If a clinic uses a scanner, it should be an integrated part of the consultation process and fee, not an additional revenue stream,“ believes Tan. “Charging patients extra for automated data that still requires expert human interpretation, does not deliver genuine clinical value.”

Not only that, there should be a clear distinction made by clinics regarding what kind of scanner they are using. One that simply offers a printed report of scores and automated product recommendations is not worth paying for, especially since the data alone can be somewhat unethical to simply hand over.

“When a scanner hands a 28-year-old an ‘internal skin age’ of 42 based on arbitrary metrics or normal freckling, it can be quite alarming. Used responsibly, the report can be educational; used commercially, it can become a very persuasive sales tool. Skin is not a maths problem to be solved with a score,” says Mihalache.

Ultimately a face scan is a tool, not a service in itself. They are useful especially for those beginning a long-term programme for acne, pigmentation or sun damage, as a baseline scan can offer objective markers to track against over time.

But their true value is entirely determined by what the practitioner does with it, and as El Muntasar puts it, “a good consultation without a scanner is far more valuable than a poor consultation with one.”