Research & principles
A more thoughtful approach to AI in health.
Good questions come before confident claims. These primary sources offer a starting point for understanding AI in health: the people affected, the limits of the technology, and the responsibilities that come with using it.
Reading room
Three primary places to begin.
The sources below are independent references. They do not establish VibeCheck.luxury’s clinical effectiveness, current feature availability, compliance posture, or security controls.
- WHO · 2021
Put people and rights first.
WHO’s ethics and governance guidance places ethics and human rights at the center of AI design, deployment, and use in health. It considers responsibilities for the organizations building these systems and the people affected by them.
Read WHO’s ethics and governance guidance Published June 28, 2021.
- WHO · 2025
Understand what generative AI can—and cannot—establish.
WHO’s guidance on large multi-modal models introduces systems that accept different kinds of input and generate different forms of output. Its overview cautions that broad capability across tasks is not yet established; a convincing response still needs appropriate evaluation.
Read WHO’s guidance on large multi-modal models Published March 25, 2025.
- NIST · 2023 / 2024
Treat risk management as ongoing work.
NIST’s voluntary AI Risk Management Framework addresses risks to people, organizations, and society across design, development, use, and evaluation. Its companion Generative AI Profile focuses on risks specific to generative systems.
Explore NIST’s AI Risk Management Framework Framework released January 26, 2023; companion profile released July 26, 2024. NIST states AI RMF 1.0 is being revised.
Editorial synthesis
Questions worth asking.
These are VibeCheck.luxury’s suggested questions for a conversation about AI in a practice. They are not quotations from the sources above or a formal assessment instrument.
- What information does the tool receive, and who can access it?
- Which decisions remain with a person?
- What evidence supports the specific task being offered?
- How are mistakes, corrections, and limits communicated?
VibeCheck.luxury in beta
Explore the intended product shape.
Research references do not validate a product. Learn what VibeCheck.luxury is being developed to be, then confirm current beta details directly.