One synthetic day (2026-09-12) at Lakeview General Hospital. Everything below is computed from the decision stream: nothing is a typed-in number.
Decisions that need human attention first. Held decisions and who must decide — nothing executes while a hold is open.
0
Who may do what without a human in the loop.
The agent drafts, summarizes, and suggests. A human performs every action that changes a record or leaves the boundary.
The agent performs bounded internal actions inside its envelope. Anything with external effect, elevated risk, or irreversibility is held for a human.
The agent runs a workflow end to end inside a tightly specified envelope. Humans set direction, receive receipts, and check in on exceptions.
How Nexus Intelligence and Nexus Readiness change what Alice enforces. Fact → inference → adjustment.
Assessment v2 (49/100, Foundational) has 3 critical blockers: GOV-04, DATA-05, VEN-02.
Shadow AI is present (GOV-04), so any agent attached to a discovered system cannot be trusted at Level 2 or 3 until the inventory is verified.
Policy DEMO-POL-006 caps such agents at Level 1 and reviews every non-public request.
Ontario Regulation 51/26 is in force: prescribed hospitals must operate a documented cyber-security program, run maturity assessments, and report incidents on prescribed timelines.
Lakeview is a prescribed entity. AI systems in clinical workflows sit inside the program's scope, so the inventory (GOV-01), incident reporting (SEC-03), and vendor oversight (VEN-01) controls now carry a binding deadline rather than a best practice.
Draft policy DEMO-POL-005 puts external tool execution inside incident-reporting scope; applied as REVIEW pending committee approval.
7 of 11 records in the current filter involve a third party (64%).
Where PHI or health data is processed outside the organization's boundary, the custodian's accountability does not travel with the data. Vendor concentration (clearinghouses, file-transfer tools, shared-services organizations) turns one breach into many.
| Policy | Status | Decisions |
|---|---|---|
| Data Sovereignty Policyv1.0 · approved policy | approved | 1 |
| External Publication Policyv1.0 · approved policy | approved | 2 |
| Incident Reporting Alignment (O. Reg. 51/26)v0.9 · intelligence signal | draft | 2 |
| Autonomy Cap for Unverified Inventory (GOV-04 gap)v1.0 · readiness gap | approved | 2 |
Alice on this site runs over the synthetic Lakeview General Hospital environment. Verdicts, checks, and receipts are computed by the Nexus decision engine on synthetic requests; nothing here is a customer, a patient, or a production deployment. Signed-in organizations see their own agents, decisions, and receipts.