"AI doctor" chats and symptom checkers
A ChatGPT-style chat that answers "what do I have?" with a diagnosis. Reviewers test it with a few symptoms and read the answers.
What 1.4.1 says and how the message reads
Guideline 1.4.1 says medical apps that could provide inaccurate data or that could be used to diagnose or treat patients may get greater scrutiny. Apps must disclose data and methodology behind accuracy claims for health measurements, and Apple rejects them if accuracy cannot be validated; measuring blood pressure, body temperature, blood glucose or blood oxygen, or taking x-rays, with the device's sensors alone is named as not permitted. Apps should remind users to check with a doctor before making medical decisions, and an app with regulatory clearance should include a link to that documentation in the submission.
The rejection cites "1.4.1 - Safety - Physical Harm" and usually says what is missing: sources for medical information, validation of a measurement, a doctor reminder, or proof of clearance. It can arrive together with 5.1.1(ix) when a healthcare service comes from an individual account, and with 1.2 when an AI chat produces the health content.
The rules a health app is measured against
- 1.4.1
Medical accuracy and doctor reminder
Claims about health measurements need disclosed data and methodology; device-only measurements of blood pressure, glucose, oxygen or temperature are not allowed; remind users to consult a doctor; link regulatory clearance if you have it.
- 1.4.2
Drug dosage calculators
Only from a drug manufacturer, hospital, university, health insurer, pharmacy or other approved entity, or with approval from the FDA or a counterpart regulator.
- 5.1.1(ix)
Regulated fields need a legal entity
Apps that provide services in highly regulated fields such as healthcare should be submitted by the legal entity that provides the service, not by an individual developer.
- 5.1.3
Health data
No use of health, fitness or medical data for advertising or data mining, no false data written to HealthKit, no personal health information in iCloud, consent and ethics approval for research.
- 5.1.2(i)
Sharing with third-party AI
Since the November 2025 update, apps must clearly disclose when personal data goes to third parties, including third-party AI, and get explicit permission first.
Where vibe-coded health apps get caught
Measurements the phone cannot make
AI tools happily generate a "measure your blood oxygen with the camera" feature. Guideline 1.4.1 names this exact case.
Dose and medication calculators
Insulin, dosage or supplement calculators from an individual developer run into 1.4.2.
Telehealth from a personal account
Booking and chatting with clinicians is a regulated service; Apple expects the provider's company account under 5.1.1(ix).
Cure claims in the listing
"Reverse diabetes", "heal anxiety in 7 days": claims in the description or screenshots are reviewed as strictly as the app.
Mental health chat without a safety net
No crisis handling, no pointer to emergency help, no way to report a harmful answer.
With a good optimized version of your web app you can have a stunning mobile app within a couple of minutes. Additionally to that the customer service is very nice and very helpful! I can just recommend WebViewGold :-)
Wellness app or medical device? A practical sorting
Regulators decide by the intended use you describe, in the app and in the listing. This table is orientation, not legal advice.
| Your app... | Usually treated as | What to prepare |
|---|---|---|
| logs habits, water, sleep, workouts or mood for general wellbeing | Wellness | Honest copy, doctor reminder where advice appears, privacy policy, "No" in the medical device declaration |
| explains conditions and treatments with cited sources | Health information | Named sources, a review date, a doctor reminder, no personal diagnosis |
| suggests a diagnosis, triages symptoms, interprets readings or calculates doses | Possibly a regulated medical device (FDA, EU MDR, UK rules) | Regulatory assessment, clearance or CE/UKCA documents, a company account, the device declaration |
| connects patients with licensed clinicians | Regulated healthcare service | Submission from the provider's legal entity, licenses, strict privacy setup |
| runs a health study | Health research | Informed consent flow and ethics board approval (5.1.3) |
Fix steps after a 1.4.1 rejection
Inventory every health claim
Listing, screenshots, onboarding, buttons, push texts and the answers your AI gives. Paste them into one document; you will be surprised what the builder wrote.
Remove or substantiate
Delete measurement claims the phone cannot make and cure promises. For information content, add a Sources page with named guidelines or studies and a "last reviewed" date.
Add the doctor reminder where advice appears
Onboarding, under every AI answer and in the description. Your web app is loaded live by WebViewGold, so wording changes need no new build.
Put guardrails on the AI
A system prompt that refuses diagnoses and doses, sends emergencies to local emergency services, and a report button for harmful answers (see Guideline 1.2 for AI content).
Handle data like health data
Name the AI provider in the privacy policy, ask for explicit consent before health details go to it, never use health data for ads. With the HealthKit plugin (
readhealthkit://,writehealthkit://), add both purpose strings and request only the types you use (HealthKit docs).Sort out the seller and the paperwork
Regulated service or device: submit from the company account, attach clearance or CE/UKCA documents and their links in the review notes, and complete the medical device declaration.
Reply with evidence
List what changed, where the reminder and sources are, and attach documents. Use the template below.
Guardrails for an AI health or wellness chat
Add this to the system prompt of your OpenAI, Anthropic or Gemini call. It does not make a medical device safe, but it stops the answers reviewers reject most. More on AI apps in general: AI chat app store rules.
System prompt block to add to your AI chat (adapt to your app)
You are a wellness assistant in [app name]. You give general information about [sleep / nutrition / fitness / stress], not medical care. Rules: - Never diagnose, never name a likely condition for the user's symptoms, never interpret lab values or readings as normal or abnormal. - Never recommend, calculate or change doses of any medication or supplement. - If the user describes chest pain, trouble breathing, signs of stroke, severe bleeding, thoughts of self-harm or any emergency: tell them to contact local emergency services now and stop giving other advice. - End every answer that touches health with: "This is general information, not medical advice. Please check with a doctor before making decisions about your health." - When you state a fact about health, name the type of source (for example national health guidelines) and do not invent studies or numbers. - Do not ask for more health details than needed, and never ask for names, addresses or insurance numbers.
Reply template for Resolution Center
Resolution Center reply (adapt the bracketed parts)
Hello App Review team, Thank you for the review under Guideline 1.4.1. We have made the following changes in build [number] and in our metadata: 1. Claims: we removed [measurement / cure claims] from the app and the listing. The app is a [wellness / health information] app and does not diagnose or treat. 2. Sources: health information now cites [sources] on a Sources page ([where in the app]). 3. Doctor reminder: shown [during onboarding, below every assistant answer and in the description]. 4. AI safeguards: the assistant declines diagnoses and dosing, refers emergencies to local emergency services, and users can report any answer. 5. Data: health details are only sent to [AI provider] after explicit consent; this is described in our privacy policy at [URL]. No health data is used for advertising. [6. Regulatory status: the app is cleared as [device class / CE / UKCA]; documentation: [link]. It is submitted by [legal entity], the provider of the service.] Best regards, [Name], [Company]
Pre-submission checklist for health apps
- No measurement the device cannot validate, no cure or treatment promise anywhere in app or listing.
- A doctor reminder wherever the app gives advice, visible without scrolling through legal text.
- Sources for medical information, with a date.
- AI answers tested with 20 tricky questions (dosing, emergencies, self-harm) before submission.
- Consent before health data goes to an AI provider; privacy policy names the provider (see privacy labels guide).
- Category, age rating answers and the medical device declaration consistent with each other.
- Regulated service or device: company account, documents linked in the review notes.
Frequently asked questions
Can I publish an AI health chatbot as an individual developer?
A general wellness or education assistant with strong guardrails can be published from an individual account. Once the app diagnoses, treats, calculates doses or provides a healthcare service, expect Apple to ask for a legal entity under 5.1.1(ix) and for regulatory documents.
Is a disclaimer enough to pass 1.4.1?
No. A doctor reminder is required, but it does not cancel a claim the app cannot support. Remove the claim, then add the reminder.
My app is a fitness tracker. Does the medical device declaration apply to me?
If your category is Health & Fitness and you distribute in the EEA, the UK or the US, you answer the question; for a fitness or habit tracker the answer is normally No, which takes a minute.
Can a WebViewGold app read Apple Health data?
Yes, through the HealthKit plugin, which is activated with a plugin token from WebViewGold and the HealthKit capability. Your web app reads values with readhealthkit:// and writes with writehealthkit://, and the Guideline 5.1.3 rules on health data apply.
Should I choose the Medical or the Health & Fitness category?
Choose by what the app does. Medical suits apps for patients or clinicians; Health & Fitness suits workouts, nutrition, sleep and wellbeing. Either category triggers the medical device question in the regions listed above.
Three ways to get your app into the stores
Same result, different amount of work on your side. All three come with the approval guarantee and 24/7 human support.
Do it yourself
$109 per platform, one-time
The WebViewGold license: a finished Xcode or Android Studio project. Paste your URL, set name and icon, switch on features, build and submit. Free updates forever.
See plans on the main pageCloud Builder
$349 per license, one-time add-on
No Mac, no Xcode, no Android Studio: configure in the browser, we build both apps in the cloud and hand them to App Store Connect and the Play Console.
How the Cloud Builder works Try the free demoFull Service
$698 per platform $798
Send us your URL and icon. Our team configures, builds and delivers your app within 72 hours, white-label, with source projects. Both platforms for $1,158.
Full Service details Powered by
Winter Sale prices valid until December 31, 2026.