AI agents are already scheduling appointments, running prior authorization and touching patient records. HIPAA requires you to know who accessed PHI, and when the requester is an agent, most systems cannot answer that. This on-demand session is built for the people who sign off together: privacy and compliance on one side, engineering and product on the other.
Part of Agentic Identity, Explained, a five-part webinar series from Vouched. We wrote KYA-OS, the open standard for agent identity and delegation, and donated it to the Decentralized Identity Foundation. Each session shows the standard working in a real environment.
Key takeaways
- Answering "who authorized this agent to access PHI" with evidence, not logs
- Capturing and scoping consent for both patient agents and clinician agents
- Building an audit trail that survives regulatory review, with proofs any third party can verify
- Explaining a decision when the model has changed three times since it was made
- How verifiable delegation lets clinical teams move faster by making guardrails provable
- What is safe to put in a prompt in a regulated environment, and who gets to decide
Speakers
Dylan Hobbs
Founding Principal Engineer, Vouched
Rosalyn Curato
Chief Innovation Officer & GM, Agentic Security, Vouched
Brigitte Piniewski, MD
Founding Network Lead, BlockHaven Inc.