ScanSkinAI Medical Review Policy

Content is drafted by the ScanSkinAI editorial team from published clinical guidance, peer-reviewed literature and public-health sources. Drafts describe what a condition looks like, what warning signs are recognised in the literature and when to seek professional care. They never provide individual medical advice.

The same standards apply across the site — from the A–Z skin condition library and the ScanSkinAI blog to triage guidance shown inside the product. Every source we rely on is listed on our sources & references page.

Clinical content — skin cancer, moles and pigmented lesions, symptom identification, rashes, wounds and triage guidance — is prioritised for review by a clinician on our clinical review team. Non-clinical content, such as product explainers or company updates, is edited but not medically reviewed, and is not labelled as reviewed.

Where an article carries a "Medically reviewed by" line, a named clinician has read that article and confirmed its clinical statements. Where no line appears, the article has not been through medical review.

• Medically reviewed by [name], [date] — a named registered doctor has read the page and confirmed its clinical statements as of that date.
• Written by the ScanSkinAI editorial team — sourced and edited, but not yet through clinical review.
• AI-generated result — produced by our model from your photo and answers. No clinician has seen it.
• Dermatologist reviewed case — a clinician has personally assessed your submitted photos through the paid dermatologist review service.

Reviewers are qualified medical doctors, in most cases dermatologists, holding current registration with a recognised medical regulator. We record each reviewer's credentials, registration body and areas of expertise on their profile page, and we publish only credentials the clinician has supplied and we can attribute to them.

12 clinicians are currently published on our clinical review team, spanning dermatology, general practice, paediatric and inflammatory skin disease. Qualified dermatologists can apply to join through become an online doctor.

We prefer national and international clinical guidelines, systematic reviews and peer-reviewed primary research, followed by major public-health and specialist bodies. Single-study claims, preprints and commercial material are not used as the sole basis for a clinical statement.

Claims about how our own technology performs are held to the same standard and are separated from general medical information. Our validation data, methodology and limitations are published on the evidence base, the accuracy report and in our research hub.

Clinical articles are scheduled for re-review at least every 24 months, and sooner when guidance changes, when a reviewer flags an issue, or when a reader reports a problem. The date shown next to a reviewer's name is the date of the most recent medical review.

When guidance changes, the affected articles are re-drafted, re-reviewed and re-dated. Where a change is material to a reader's decisions, the update is described in the article rather than made silently.

Suspected errors can be reported to info@ScanSkinAI.com. Reports affecting clinical safety are prioritised, assessed with a reviewer, and the article is corrected or withdrawn.

ScanSkinAI operates several distinct things, and we keep them separate:

• AI screening — an automated assessment of a photo, intended for screening, education and decision support. It is not a diagnosis.
• AI-generated information — explanatory text produced alongside a result. It is general information, not personal advice.
• Medical content review — a clinician checking the accuracy of an article on this website.
• Dermatologist review — a paid service in which a clinician reviews your submitted photos and returns written feedback.
• Clinical diagnosis — something only an in-person or appropriately regulated clinical assessment can provide.

Not every AI result is reviewed by a dermatologist. Human review happens only when you request our dermatologist review service.

We state these openly because a screening tool that is used without understanding its limits can cause harm:

• A result depends heavily on photo quality — lighting, focus and distance all change what the model can see.
• Lesions that are very small, obscured by hair, or on skin tones and body sites that are underrepresented in public dermatology datasets are harder to assess.
• The model assesses appearance only. It has no access to your history, family history, symptoms over time, or how a lesion feels.
• A low-risk result never rules out disease. Any lesion that is changing, bleeding, itching or simply worrying you should be seen in person, whatever the app says.

How we measure and report performance, including where the model is weakest, is documented in the accuracy report. Data handling is covered in our privacy policy and terms of service.

Reviewers disclose relationships that could reasonably influence their assessment and do not review content covering a product or service in which they hold a material interest. Commercial partners do not commission, approve or amend clinical content.

Clinicians on the ScanSkinAI clinical review network are independent professionals, not employees. Their clinical judgement is their own, and editorial or commercial teams cannot overrule it. A reviewer may decline to approve content.

ScanSkinAI does not provide emergency care and cannot examine you physically. If you have a rapidly changing lesion, bleeding, a spreading infection, severe pain, breathing difficulty or any other urgent symptom, contact your local emergency service or attend a healthcare facility immediately rather than waiting for any online result.