How SkinVault is meant to be used
Effective: 4 May 2026 · Version 1.0
Read alongside our Terms of Service and Privacy Policy.
⚠ Read this first
SkinVault is clinical decision support software. It is not a medical device. AI-generated outputs are assistive suggestions only and do not constitute a diagnosis, prescription, or treatment plan. The treating clinician retains full clinical responsibility for every decision.
1. What SkinVault is
SkinVault is a record-keeping and workflow tool for registered medical practitioners managing dermatological and skin cancer cases. It supports clinicians by:
- Storing patient demographics, lesion records, procedure logs, and histopathology results.
- Capturing and organising clinical and dermoscopic images.
- Generating audit reports aligned with the Skin Cancer Audit Research Database (SCARD) format.
- Producing AI-assisted differential lists, dermoscopic descriptions, surgical plans, and referral letters when the clinician requests them.
- Transcribing clinician dictation into structured clinical text.
Every feature is designed to save administrative time, not to replace clinical reasoning.
2. What SkinVault is not
- SkinVault is not registered as a medical device with the Therapeutic Goods Administration (TGA) in Australia, nor with Medsafe in New Zealand.
- It is not a diagnostic device and is not intended to diagnose, treat, cure, or prevent any disease.
- It is not a substitute for histopathological examination, specialist clinical assessment, or independent medical judgement.
- It is not intended for use by patients directly. SkinVault must only be used by registered clinicians and authorised practice staff.
3. Limits of AI outputs
The AI features in SkinVault rely on large language and vision models supplied by third parties. These models:
- May produce confidently incorrect outputs (commonly called “hallucinations”).
- May miss clinically significant findings, especially in atypical or amelanotic lesions.
- Reflect biases in the data they were trained on, and have not been validated in randomised clinical trials for dermatological diagnosis.
- Cannot incorporate information that is not visible in the supplied image or stated in the supplied clinical context.
- Should never be the sole basis for a decision to biopsy, excise, monitor, or discharge a patient.
Confidence scores presented by SkinVault are calibrated estimates only. They do not represent clinical probability of disease and must not be interpreted as such.
4. Clinician responsibility
By using SkinVault, you confirm that you:
- Are a registered medical practitioner or nurse practitioner with current registration in Australia or New Zealand.
- Will exercise independent clinical judgement on every patient.
- Will not act on an AI output without first verifying it against the patient's history, examination findings, and (where indicated) histopathology.
- Will not include AI output in a discharge summary, referral, or patient communication without first reviewing and editing it for accuracy.
- Are responsible for obtaining valid patient consent for clinical photography and for storing those images in SkinVault.
5. No emergency or urgent care
SkinVault is not designed for emergency, urgent, or time-critical clinical scenarios. Do not rely on SkinVault for any decision that requires same-visit or same-day action. If a patient is unwell, refer them to an appropriate emergency service.
6. Reporting issues
If an AI output appears materially incorrect, dangerous, or inconsistent with clinical findings — or if you suspect a software defect that could affect patient safety — please report it promptly:
Email: safety@skinvault.app
Subject line: [SAFETY] <short description>
We aim to acknowledge safety reports within one business day and investigate promptly. We will notify the broader user base where a defect has the potential to affect clinical decision-making at scale.
7. Changes to this disclaimer
We will update this disclaimer as our AI features evolve. The version date at the top reflects the most recent revision.