AI Supplemental Terms
Version 1.0 · Effective 2026-05-18
These terms govern AI-powered features across Helix, LedgerOne, and Volt. They supplement the Orvica General Terms of Service. AI-related claims and limitations apply whenever you use any feature marked as AI-powered.
1. AI limitations
AI-generated outputs may be inaccurate, incomplete, misleading, outdated, biased, or inappropriate for your circumstances. You are responsible for independently verifying all AI-generated outputs before relying on them.
2. No professional advice
AI features do not provide medical, legal, financial, tax, investment, emergency, or professional advice of any kind. AI outputs are informational only and are not a substitute for consultation with a qualified professional.
3. No sole reliance
You must not rely solely on AI-generated outputs to make medical, financial, legal, safety, or emergency decisions.
4. Human review required for health outputs
For health-related outputs, including medication information, pill identification, drug-interaction hints, symptom summaries, appointment notes, and caregiver summaries, you must consult a licensed healthcare professional or pharmacist before acting.
5. AI training of your content
We do not use Your Content to train third-party foundation models unless we obtain your express consent or state otherwise in a product-specific agreement. Where AI providers offer the option, we contractually restrict them from using Your Content for model training.
6. AI provider disclosure
AI features are powered by third-party model providers. Current providers:
- Anthropic (Claude models) — used for summarization, document parsing, and rebuttal-letter drafting. Sensitive-data features are gated until a Business Associate Agreement is in effect for PHI workloads.
7. In-product acknowledgements
Certain AI features (pill identification, medication summarization) require an in-product affirmative acknowledgement that you understand the output is not medical advice and must be verified. We log these acknowledgements with timestamp and user ID for compliance recordkeeping.
8. Errors and reporting
If you believe an AI feature has produced harmful, false, or otherwise problematic output, please report it to hello@orvica.co. Orvica will investigate and may disable the affected feature pending review.