# Five Questions for Responsible Health AI Leadership

Source: https://mission.convergenceq45.org/blog/five-questions-responsible-health-ai

Publisher: Convergence Q45

Author: Convergence Q45

Published: 2026-11-07

## What should a health AI evidence brief contain?

Describe purpose, supporting evidence, evaluation across relevant groups, responsibility for concerns and updates, and how people can understand and participate in the service.

## Begin with the people being served

Responsible health AI leadership begins by naming the people, task, and setting involved. A useful tool addresses a specific need. Leaders can make that need visible by writing a one sentence purpose statement and inviting the people affected to help refine it. This Convergence Q45 framework translates the themes of stewardship and service into questions an organization can use before adopting a technology.

## Ask about purpose and evidence

Ask the developer to describe the intended user, supported task, and expected action. Request the study population, evaluation setting, comparison method, and known limitations. For a clinical application, involve the relevant clinical and regulatory professionals in interpreting that evidence. FDA guidance on clinical decision support and predetermined change control plans offers useful source material for understanding software functions and planned device updates.

## Examine equity and accountability

Ask how performance was evaluated across relevant groups and how uncertainty is communicated. Explore practical access questions: language, disability access, device availability, connectivity, and the time needed to use the service. Then name a responsible owner. Record how concerns are reviewed, how changes are communicated, and what triggers another assessment. These are editorial evaluation questions; the appropriate evidence and regulatory pathway depend on the product and intended use.

## Keep human dignity visible

Invite people to explain the service in their own words. Can they identify its purpose, ask a question, and reach a responsible person? Prepare a short evidence brief with five fields: purpose, evidence, equity, accountability, and human dignity. Give each field an owner, a source, and a next action. The resulting record can support thoughtful decisions and continued learning as the technology changes.

## Sources and further reading

- FDA · Clinical decision support software: https://www.fda.gov/regulatory-information/search-fda-guidance-documents/clinical-decision-support-software

- FDA · Predetermined change control plans for AI enabled devices: https://www.fda.gov/regulatory-information/search-fda-guidance-documents/marketing-submission-recommendations-predetermined-change-control-plan-artificial-intelligence

## Common questions

### What should a health AI evidence brief contain?

Describe purpose, supporting evidence, evaluation across relevant groups, responsibility for concerns and updates, and how people can understand and participate in the service.

### Who should review the brief?

Include the relevant clinical, technical, regulatory, and community perspectives for the intended use. Assign a responsible owner to coordinate the review.

Continue with an open guided practice: https://mission.convergenceq45.org/blog/responsible-innovation