Home · How We Use AI

How We Use AI

Version 1.0 · Effective August 15, 2026

A licensed physician makes every clinical decision at Conduit. AI never diagnoses you, never prescribes, never sets or changes your dose, and never decides whether you are approved for treatment.

1. What AI never does

  • It does not decide whether you are approved. A board-certified physician reviews your intake and makes that call.
  • It does not prescribe, and it does not set or change your dose. Only your physician does.
  • It does not replace your care team. Anything clinical is answered by a person.
  • It does not act on your behalf or take actions in your account.

2. Where we use AI today

Conduit does not currently use AI in any member-facing feature. No AI system reads your intake, participates in your care, or communicates with you.

We are building two things, and we will update this page and tell members before either becomes active: a Companion that answers practical questions about using your protocol and the portal, and an internal Evidence Engine that helps our team summarise published research for the guides we write. Neither will make clinical decisions.

3. What data AI would see

When these features become active, we will send the minimum information a feature needs to work, and we will state on this page exactly what that is before it happens. We will not sell your information to an AI provider, and we will not permit your health information to be used to train third-party AI models.

Where health information regulated under health-privacy law is involved, we will only use providers operating under the contracts those laws require.

4. Human oversight

Every AI feature we build is designed with a person accountable for the result. Clinical questions route to your care team. Content that AI helps produce is reviewed before publication. If an AI feature ever gives you an answer that seems wrong, tell us through our Contact page — a person reviews every report.

5. Your choices

You will be able to use Conduit without using any AI feature. Choosing not to use one will not affect your care, your eligibility, or your pricing.

6. What AI gets wrong

AI systems can be confidently wrong. Any assistant we offer may give an answer that is incomplete or inaccurate, and you should not rely on it for medical decisions. Your physician and care team are the authority on your treatment — if an assistant and your care team ever disagree, your care team is correct.

7. Changes

We will update this page whenever our use of AI changes materially, and note the effective date. For significant changes we will tell members directly rather than relying on this page alone.

See also our Consumer Health Data Privacy Policy, Privacy Policy, and Telehealth Informed Consent.