Session
When Workforce AI Touches Health Data
Governance beyond the HR system
PAWorld26 San Francisco · 23–24 September 2026 · San Francisco
← Back to agenda
– · Work Design
About this session
Workforce-AI governance becomes more complex when tools, browser layers or enterprise integrations cross from HR systems into environments containing PII, PHI or regulated clinical information. Sandy Pham will use contrasting worked scenarios to examine data classification, permissions, shadow tools, vendor and business-associate obligations, accountable ownership, human review, monitoring and stop authority. Delegates will leave with a practical boundary checklist rather than a claim that every control described has already been implemented at Kaiser Permanente.
Learning outcomes
- Recognise when a workforce-AI use case crosses a meaningful data or system boundary.
- Apply a checklist covering permissions, vendors, human review and stop authority.
- Decide when additional controls are sufficient and when the use case should be redesigned or stopped.
This session will explore
- How an ordinary HR use case can cross into a materially different regulated-data environment.
- PII, PHI, clinical-data, permission and system-boundary questions.
- Two worked scenarios covering lower-risk and higher-risk workforce-AI uses.
- Vendor obligations, browser-layer tools, shadow AI and accountable human review.
- A practical decision sequence for approving, redesigning or stopping the use case.
Context
A Day 2 Decision Room / Studio technical and foundation session.
Speakers and contributors
Sandra PhamPresenterVice President, Finance, People, & Business AnalyticsKaiser Permanente