liveBook · Manning

Fallback

As explained in AI Governance

Fallback logic is a programmed alternative behavior invoked when a primary operation fails, is unavailable, or cannot safely proceed.

Fallback logic sends a case to a clinician when a model result is low-confidence or unclear, or when no approved patient-facing message fits. It prevents Derm-Cam from improvising a diagnosis or reassurance. section 3.3.3

Derm-Cam escalation

Low-confidence or unclear image results go to a clinician. Derm-Cam can select only from approved patient-facing messages; if none applies, the case goes to a clinician. section 3.3.3

How AI Governance covers it

How deep it goes
Defines low-confidence, unclear-result, and unavailable-approved-message triggers for clinician escalation.
Examples
Uses Derm-Cam image results and approved patient messages to illustrate when clinician review takes over.
What it assumes
Requires clinician availability and a small set of approved patient-facing messages.
What you can do after
Configure clinician escalation for low-confidence, unclear, or unhandled outputs.
How it connects
Links model confidence and approved patient-facing messages to clinician review.

Also explained in

Compare how 24 books explain fallback →

Related concepts

This explainer was generated by Manning's knowledge graph from the text of AI Governance. The section links point to the parts of the book it draws on.