Someone notices a change on their skin. A new mole, a rash that won’t resolve, a mark that looks slightly different from before. The immediate need is simple: should I worry about this?
Yet for most people, there is no clear first step. They search online, compare images, or ask friends for reassurance, none of which provides structured medical guidance. The healthcare system, meanwhile, typically begins only once a patient formally enters it through a referral or specialist appointment.
Between noticing a concern and receiving reliable guidance lies a gap that dermatological care pathways have largely overlooked.
The scale of this gap is structural, not just a handful of isolated stories. In the United Kingdom, dermatology referral-to-treatment waits range from 7 to 17 weeks, with only 64% of patients seen within 18 weeks (NHS England, 2024). In Belgium, waiting times approach four months (Lambert et al., 2024). These figures capture only the specialist queue, a far larger population never enters the system at all during the early phase of concern.
Skin diseases affect more than two billion people globally. In 2021, they accounted for 4.69 billion new cases and ranked as the fourth leading cause of non-fatal disease burden worldwide (GBD, 2021). In 2025, the World Health Organization formally recognised skin disease as a global public health priority.
Urban, digitally connected adults face friction rather than ignorance, long waiting times, mandatory GP referrals, and competing responsibilities. They already use digital pharmacies, telemedicine, and health apps, yet these platforms have only recently begun offering a structured first step for skin concerns.
Rural and suburban populations face a different problem. Dermatologist availability drops sharply across rural Germany, southern and eastern Europe, and much of the UK. General practitioners, who often serve as the only accessible point of care, show diagnostic agreement with dermatologists of roughly 57–65% (Escalé-Besa et al., 2023). For these populations, AI-assisted triage embedded in accessible digital platforms may represent the only structured initial contact point.
Older screening cohorts carry elevated melanoma risk and often hold statutory screening entitlements. But entitlement does not guarantee navigation. Simpler entry points, pharmacy-based kiosks or insurer-linked applications, can convert passive eligibility into active access.
Lower-income groups face the widest gap: longer public system waits, limited private alternatives, and compounding barriers like transport costs and lost wages. Embedding triage within widely used digital platforms lowers barriers precisely where the structural gap is most acute.
AI in dermatology currently operates across three distinct tiers, each serving different points in the care pathway.
Autonomous diagnostic AI provides skin cancer triage directly in clinical settings. These systems are CE-marked Class III under EU MDR, designed for high-risk clinical decision-making but limited to narrow condition sets. They require specialist infrastructure and operate within formal care pathways.
Professional diagnostic support tools , typically Class IIa, assist clinicians in interpreting lesions. They improve workflow efficiency but require trained operators and cannot be embedded into consumer platforms.
Informational triage infrastructure offers broader condition coverage and can be both consumer-facing and platform-integrated. Tools in this tier that process medical images typically fall under EU MDR requirements, with the applicable class depending on intended use and risk. They do not provide a diagnosis; they organise early concerns and direct users toward appropriate professional follow-up.
The first two tiers play important roles at the specialist level. Neither addresses the stage before consultation, when an individual has a concern but lacks a structured way to evaluate it. That is the domain of the third tier.
What informational triage actually does
The platform ecosystem journey
Informational triage functions as a structured first step inside platforms people already use. A person who notices a skin change begins within a familiar environment, a pharmacy service, a telemedicine portal, a health app. An embedded AI skin-check tool allows the user to upload an image and answer a few questions. The system then provides informational suggestions about possible conditions and next steps.
When escalation is needed, the transition happens within the same platform. The user is routed directly to a teleconsultation, a pharmacist review, or a GP booking: no additional apps, no referral paperwork, no weeks of waiting simply to determine whether the concern warrants attention.
Because the AI has already organised the initial information, the professional enters the interaction with a clearer starting point. The conversation can focus more quickly on clinical judgement and next steps.
What the AI does and doesn’t
The AI functions as an informational layer, not a diagnostic authority . It analyses user-provided images and structured responses to generate ranked informational suggestions. It does not issue medical diagnoses, prescribe treatments, determine clinical urgency, or substitute for professional judgement.
Responsible deployment means operating within established regulatory and data-protection frameworks: CE marking under EU MDR, GDPR compliance, and secure EU-based hosting with anonymised data processing.
In the United Kingdom, AI-supported dermatology tools are being evaluated and deployed within NHS pathways. Teledermatology initiatives allowing GPs to transmit high-resolution images to dermatologists have resolved approximately 64% of cases within primary care, with around 90% of queries answered within 36 hours.
In the Netherlands, a large population-based deployment offered AI skin cancer assessment to approximately 2.2 million adults through a Dutch health insurer. Users reported 30% more (pre)malignant lesion claims than controls, indicating earlier detection potential, though also higher dermatology-related costs, highlighting the importance of professional escalation pathways to manage utilisation (Smak Gregoor et al., 2023).
In Germany, consumer-facing digital dermatology assessment tools are available within regulated markets, reflecting a growing ecosystem of AI-supported skin health services integrated into digital health platforms.
In São Paulo (2021), a teledermatology deployment involving 30,976 patients found that 53% of dermatological cases were managed within primary care, resulting in a 78% reduction in in-person specialist waiting times.
A broader meta-analysis comparing AI with clinicians across 19 studies reported AI sensitivity of 87% versus 80% for clinicians in skin cancer detection (Salinas et al., 2024). And a 2024 meta-analysis across 12 studies and 67,000+ evaluations found that AI support improved clinician sensitivity from 75% to 81% and specificity from 82% to 86% (Krakowski et al., 2024, npj Digital Medicine).
Compressed time-to-information. In systems where the gap between recognising a concern and the first professional consultation spans weeks or months, informational triage compresses that interval from the moment of concern.
Democratised access. Individuals in rural areas, on extended public waiting lists, or unable to afford private consultations can access the same quality of preliminary guidance as those in well-served urban centres.
Better-prepared patients. When patients subsequently consult professionals, they arrive with clearer concerns and documented visual information, enabling more focused and efficient consultations.
Europe has around 60 dermatologists per million people; the global average sits at roughly one per 60,000. These disparities prompted the 2025 World Health Assembly to explicitly encourage digital health tools for expanding dermatological access.
European maturation as the development path
Embedding informational skin triage enables support from initial concern through product guidance and, when necessary, professional referral, all within the existing platform.
Telemedicine platforms benefit from improved consultation quality . Patients arrive with documented concerns and preliminary context, allowing consultations to begin more efficiently.
Consumer health apps gain a recurring engagement feature. Users periodically checking skin changes create sustained interaction, and triage can function as a freemium layer driving progression to premium services.
Insurers gain population-scale early engagement with potential health concerns. Earlier identification supports preventive care strategies, and evidence from Dutch deployments demonstrates this approach can operate effectively at scale.
It analyses smartphone images and returns a ranked list of the five most likely conditions across 70+ skin diseases in under one second.
Since 2018, Autoderm has processed over 2 million API calls across live deployments with telemedicine, pharmacy, and consumer health platforms across Europe. It is designed as B2B2C infrastructure, an API that integrates into the platforms people already use, rather than a standalone consumer app.
Autoderm’s clinical evidence portfolio spans 10 clinical studies across five countries (Sweden, Spain, China, Uganda, United Kingdom), including four independent peer-reviewed publications. Three reader studies consistently demonstrate that clinicians using Autoderm outperform clinicians working without AI assistance, with the Boots UK GP Reader Study (2024) showing a 21 percentage point improvement in GPs’ top-1 diagnostic accuracy and a 40% reduction in unnecessary specialist referrals.
Post-market safety data spans 2 million+ API calls with zero adverse events since launch.
Regulatory status
Autoderm operates at the highest regulatory standard for AI dermatology software at this classification level:
CE marked under EU MDD 93/42/EEC as a legacy Class I medical device, currently transitioning to MDR Class IIa under EU MDR 2017/745, with the technical file submission planned for 2026; Autoderm operates under Article 120 legacy provisions until that transition completes
FDA Breakthrough Device Designation, granted by the U.S. FDA, signalling FDA recognition of the clinical unmet need Autoderm addresses
No FDA-cleared AI dermatology device currently exists for smartphone-based multi-condition classification.
This breadth matters because the majority of skin concerns are not cancer. A system that can only answer “suspicious or not suspicious” leaves the vast majority of users without useful guidance. Autoderm’s multi-condition approach addresses the full spectrum of what people actually worry about when they notice a change on their skin.
The next phase is wider integration across pharmacy ecosystems, telemedicine services, and health systems globally. As these tools mature, they create a structured entry point connecting early consumer concerns with professional care pathways.
Beyond Europe, the same structural gap exists in more acute form. Smartphone ownership and AI familiarity are expanding rapidly, increasing readiness for digital health entry points. The infrastructure being validated in European and American markets will define the first point of contact for skin health globally. The question is no longer whether, but how far it reaches, and how fast.
NHS England (2024). Referral-to-treatment waiting times data. https://www.england.nhs.uk/statistics/statistical-work-areas/rtt-waiting-times/
Lambert J, Lambert J, Roegies K, Nikkels A, Garmyn M, Snauwaert J, Willaert F, Bouffioux B, Hoorens I, Vossaert K, Gutermuth J, Del Marmol V. The doctor will see you now, in 4 months: A Belgian perspective on waiting times for dermatologic care. J Eur Acad Dermatol Venereol. 2025;39(3):e204–e205 (epub 10 June 2024). https://doi.org/10.1111/jdv.20161
GBD 2021 — Global Burden of Disease Study. Institute for Health Metrics and Evaluation (IHME). https://www.healthdata.org/research-analysis/gbd
World Health Organization (2025). Seventy-eighth World Health Assembly adopts resolution WHA78.15, recognising skin diseases as a global public health priority (24 May 2025). https://www.who.int/news/item/24-05-2025-seventy-eighth-world-health-assembly—daily-update–24-may-2025
Escalé-Besa A, Yélamos O, Vidal-Alaball J, et al. (2023). Exploring the potential of artificial intelligence in improving skin lesion diagnosis in primary care. Scientific Reports 13, 4293. https://doi.org/10.1038/s41598-023-31340-1
Smak Gregoor AM, Sangers TE, Bakker LJ, et al. (2023). An artificial intelligence based app for skin cancer detection evaluated in a population based setting. npj Digital Medicine 6, 90. https://doi.org/10.1038/s41746-023-00831-w
Salinas MP, Sepúlveda J, Hidalgo L, et al. (2024). A systematic review and meta-analysis of artificial intelligence versus clinicians for skin cancer diagnosis. npj Digital Medicine 7, 125. https://doi.org/10.1038/s41746-024-01103-x
Krakowski I, Kim J, Cai ZR, et al. (2024). Human-AI interaction in skin cancer diagnosis: a systematic review and meta-analysis. npj Digital Medicine 7, 78. https://doi.org/10.1038/s41746-024-01031-w
World Bank (2025). Global Findex Database 2025. https://www.worldbank.org/en/publication/globalfindex
GSMA (2025). Mobile Internet Connectivity Report. https://www.gsma.com/r/somic/
