NIN
NIN develops the profession and the ecosystem. Community, education, workforce empowerment, implementation support, partnerships, and innovation belong here.
Public governance register · NIN–NAIO Master Directive v1.1
How Nurse AI OS is governed
One doctrine spine, clear authority boundaries, and status claims that change only when evidence changes.
Version 1.1 Effective July 25, 2026 Directive owner: Robert Domondon
Legal and compliance boundary: This is a governance framework, not legal advice. Adoption does not by itself establish legal, regulatory, clinical, security, privacy, or institutional compliance. Institutional reliance requires qualified counsel and organizational review.
The product category
The governed environment is the product
Nurse AI OS is the governed professional environment in which nursing applications, agents, models, knowledge, memory, tools, and workflows operate—so nurses and healthcare teams can expand capability without surrendering professional judgment, accountability, humanity, or agency.
It is not an AI model, chatbot, EHR, autonomous clinician, licensing board, regulator, or substitute for an institution's authority. The public Community experience is the accessible first layer, not proof that every destination capability exists today.
Canonical ecosystem statement
Distinct roles, one accountable system
NAIO
NAIO defines doctrine, standards, competence, and assurance. It is not a regulator, licensing board, or source of institutional permission.
Nurse AI OS
Nurse AI OS provides the governed professional environment. Editions and role packs organize access; purchase or download never grants authority.
Florence-X
Florence-X plans, routes, coordinates, and evaluates the work. It does not become the final decision-maker.
EDENA
EDENA determines what is permitted and enforces the controls. Today that enforcement is capability-specific; guidance alone is not mechanical enforcement.
Humans, models, and tools
Replaceable models and tools provide capabilities. Nurses and other authorized humans decide what enters practice and remain accountable.
Final command: Agents propose. Humans judge. Nurses steward.
Applicability
Three states govern every public claim
State A · Binding now
Binding now
Human accountability, dignity, anti-surveillance, patient transparency, data boundaries, prohibited practices, evidence rules, marketing discipline, and the Final Command apply immediately.
State B · Capability-dependent
Binding when capability exists
EDENA enforcement, Florence-X orchestration, registries, memory, connectors, editions, audit events, and Solution Packages may be called operational only where the capability exists and satisfies the directive.
State C · Formal activation
Activated only through a formal gate
Institutional pathways, services, credentials, Exchange, Observatory, conformance, and Sovereign programs become active only through their stated program or authorization gate.
EDENA decision model
Risk, data, and action are separate questions
| Dimension | Canonical values | Public meaning |
|---|---|---|
| Risk tier | Green · Yellow · Orange · Red-P · Red-E | Green supports low-risk learning and drafting; Yellow requires review; Orange influences consequential decisions; Red-P is prohibited; Red-E is exceptional controlled high-risk use. |
| Data class | D0 · D1 · D2 · D3 · D4 | D0 is public/synthetic/deidentified; D1 is personal professional; D2 is confidential organizational; D3 includes regulated or highly sensitive data such as PHI; D4 is restricted critical data. |
| Action mode | Observe · Draft · Recommend · Prepare Action · Act With Approval · Constrained Autonomy | Authority increases only through explicit controls and named human accountability. Unrestricted Autonomy is prohibited. |
Risk never grants authority. Action mode never lowers risk. Data sensitivity never disappears because a tool is convenient. Ambiguity narrows capability and routes the question to an accountable human.
Current public status
What exists, what is gated, and what is not launched
| Capability or program | Status | Scope and limitation |
|---|---|---|
| Browser-first Community experience | Available now | D0/D1 only; no PHI; SOUL Quiz, prompts, templates, local files, role workflows, and human review. |
| EDENA advisory materials | Available now | Review language and artifacts; they do not guarantee enforcement by ChatGPT, Claude, Hermes, or another host. |
| EDENA mechanical enforcement | Capability-gated | Only where a named implementation, control path, tested environment, and evidence exist. Not universal across the public experience. |
| Florence-X orchestration | Capability-gated | Developing and component-specific; not represented as one finished public mission-control service. |
| Personal and Professional editions | Formal-gate programs | No edition purchase or label creates employer, clinical, institutional, or data authority. |
| Institutional and Sovereign editions | Formal-gate programs | D2–D4, connected workflows, PHI eligibility, and consequential use require separate authorization and controls. |
| Education and competency credentials | Program planned / not launched | A framework exists; no public claim of licensure, specialty certification, accreditation, CE credit, or employer permission. |
| Nurse AI Exchange and Observatory | Not launched | Strategy and architecture do not establish an operating marketplace or monitoring program. |
| NAIO Conformant | Unavailable | The assurance program is disabled until its complete governance, assessor, evidence, appeal, renewal, and conflict controls exist. Conformance would not establish clinical efficacy. |
| Stewardship Council | Not constituted | Participation informs governance. Formal authority follows published decision rights. |
Claims and assurance
Evidence before adjectives
“Functional,” “ready,” “validated,” “production,” “safe,” “HIPAA compliant,” “institution-authorized,” and similar claims require a defined version, environment, intended use, evidence, assessor, date, limitations, and expiration. A prototype, interface, signature, test count, course completion, badge, subscription, or edition label does not establish clinical validity, compliance, competency, or permission.
Where a page preserves older implementation evidence, it is labeled historical and does not become the current canonical architecture merely because the artifact remains downloadable.
Public participation: Participation informs governance. Formal governance authority follows published decision rights. Until those rights and bodies exist, community input is influential—not binding institutional authority.
Start with what is available now
Use the public Community experience for personal, learning, and professional D0/D1 work. Keep patient and confidential institutional data out.
Build my SOUL file → Inspect historical implementation evidence