In the hierarchy

1. Part II: Societal & Human Concerns
2. D13. Health, Safety & Care
3. 13.1 Clinical AI

Virtues and principles to explore

Held as inquiry, not as a verdict.

• Service

  Hope

  Who is served and who benefits are not always the same.

• Dignity

  Center principle

  Every person is a protagonist, not a data point.

• Listening

  Trust

  Listen to understand, rather than listen to speak.

Starter questions

Written to probe curiosity and learning, not accusation.

1. 01Whose experience of Radiology model drift is not yet in the room — and how would we hear it?
2. 02What is already being done about this, and by whom — and how would we find out without assuming?
3. 03What conversation have we not yet had about this — and who needs to be in it?
4. 04Who is served by addressing this, and who benefits if we leave it unnamed?

Also on this branch

• Diagnostic errors20
• Triage algorithm bias25
• Care-management under-coverage16
• Sepsis-prediction reliability16
• Pathology AI edge cases16
• Ophthalmology screening error16

Frameworks and sources

• WHO guidance on ethics and governance of AI for healthframework
• EU AI Act — Annex III high-riskregulation