AI Transparency

How we use artificial intelligence and your rights under GDPR Article 22.

Our AI gives information, not medical advice. None of our AI features (Robin, cost estimates, or hospital matches) provide medical advice, diagnose a condition, or create a doctor-patient relationship. Always consult a qualified clinician about your care. We always tell you when you are interacting with AI rather than a person.

GDPR Article 22 Rights: You have the right not to be subject to a decision based solely on automated processing that produces legal or similarly significant effects. Where such processing occurs, you have the right to request human review, express your point of view, and contest the decision.

Our AI Systems

Robin, our AI assistant

Our chatbot answers your questions about procedures, destinations, and hospitals using our own data, and can help you start an inquiry. It is clearly labeled as AI, gives general information rather than medical advice, and can connect you with a person at any time. You are always told when you are talking to AI rather than a human.

Data used:

  • Your questions and the conversation
  • The procedure and destination you are exploring
  • Our own procedure, cost, and hospital data

Human review available

Hospital Matching Engine

Ranks and recommends hospitals based on your procedure, budget, location, language-support needs, and preferences using a multi-dimension algorithm. Hospital reviews are confidence-weighted (a higher review count makes a rating more reliable, but volume alone is not rewarded, so larger markets get no unfair edge). For more clinically complex cases it leans more toward well-reviewed, established hospitals. Rankings are never paid for by hospitals.

Data used:

  • Procedure requested
  • Budget range
  • Origin country
  • Timeline
  • What matters most to you (e.g. language support)
  • Case complexity (from your answers)
  • Hospital public information & reviews

Human review available

Inquiry Scoring

Scores the quality and readiness of your inquiry across 6 dimensions to help hospitals understand your needs. Scores are not shared with patients but affect which hospitals see your inquiry.

Data used:

  • Interaction history
  • Profile completeness
  • Timeline indicators
  • Budget signals
  • Medical readiness

Human review available

Cost Estimator

Generates a typical cost range for a procedure in a destination, derived from credible published reference sources and a transparent model, shown with an 'as of' year and a confidence indicator. It is our estimate, not a clinic quote. See our cost methodology for how figures are sourced and derived.

Data used:

  • Procedure type
  • Destination country
  • Origin country
  • Accommodation preference

No individual decisions

Data Gathering & Verification

We use automated crawlers to collect public information about hospitals from their own websites and public records, including hospital services, practicing doctors, and facility details. This is information about hospitals, not about you. Our team reviews key details before they are shown to patients.

Data used:

  • Public hospital websites and records
  • Hospital public information
  • Facility service details
  • Hospital contact details

No individual decisions

Content Generation

Some of our written content (educational articles, hospital and destination pages, and social posts) is drafted by AI, grounded in our own data. It is informational, never medical advice, and is checked before it is published. It does not use your personal information.

Data used:

  • Our own procedure, cost, and hospital data
  • No patient personal data

No individual decisions

How to Request Human Review

If you believe an automated decision has significantly affected you (e.g., your inquiry was not matched with clinics), you can request a human review at any time:

  • Click “Request Human Review” on any AI-generated result
  • Contact us at contact@curemeridian.com
  • Submit a support ticket from your dashboard

We aim to respond to human review requests promptly.

Fairness and Bias

We conduct quarterly fairness audits of our AI systems across protected attributes including nationality, language, gender, budget, and age. Our goal is that no group receives systematically worse outcomes. Results of fairness audits are reviewed by our Data Protection Officer. We use de-identified data to train our models and apply algorithmic guardrails to mitigate bias.

Data Used by AI

Before any text is sent to an external AI provider, we remove direct identifiers (such as your name, email, phone number, specific dates, and geographic identifiers smaller than a country) following HIPAA Safe Harbor de-identification standards. Documents and images you upload (for example scans or reports) that cannot be reliably de-identified are not sent to external AI providers at all, they are withheld by default. We continue to expand automated de-identification to more document types over time.

Automation and the EU AI Act

Most of our AI runs automatically: a person does not review every result, so you should not assume that a clinician or a staff member has checked any individual output. Where an automated decision significantly affects you, you have the right to request a human review (GDPR Article 22), using the request options listed above. We treat our chatbot as a higher-risk system under the EU AI Act: it gives general information only, never a medical or eligibility decision, and you are always told when you are interacting with AI rather than a person.

Limitations of our AI

AI can be wrong. Our systems can produce incomplete or out-of-date information, and cost estimates are our own estimates rather than clinic quotes. Treat AI output as a starting point for your own research and decisions, not as a substitute for professional medical, legal, or financial advice. We continue to test and improve these systems, but we cannot guarantee that every output is accurate.

Questions

For questions about our AI systems, contact our Data Protection Officer at contact@curemeridian.com.