Responsible AI in Health
summary
A responsible AI review board evaluates projects with a checklist — is AI needed, who are stakeholders, what are benefits and harms, does data contain PII, and could model decisions affect jobs or care — and it surfaces ethical issues so teams reformulate or mitigate risks. Practical tools used include Datasheets for datasets, a model with AUC 0.92 that still shows subgroup gaps uncovered by phalon analysis, and the interpretML explainable boosting machine for transparent feature effects; HIPAA and domain expert review guide healthcare work.
AI-written summary · about
chapters (10)
- 0:00Responsible AI in Health intro
- 0:57Historical parallels
- 3:03Biomedical ethics lessons
- 5:32Ethical AI journey
- 6:15AI review process
- 10:47Tools for Responsible AI
- 11:52Healthcare AI specifics
- 14:21Racial bias case study
- 15:05AI implementation tools
- 18:19Responsible AI resources
related talks
34mResponsible AI: The Key to Ethical, Safe and Inclusive Software Development
15mResponsible AI 2.0 : NO AI without RAI
19mA developer's guide to responsible AI review processes
33mResponsible AI in Practice: Real-World Examples and Challenges
40mSocially responsible AI: Practical tools for robust design
39mThe Search for Transparency and Accountability in the Age of...
39mResponsible AI- 39mSponsored Session: The Search for Transparency and Accountability in the Age of AI: RAI & XAI as Essential Tools