ARPA-H ADVOCATE describes a prospective clinical-agentic-AI objective

Why it matters

This moves agentic AI from administrative copilots toward clinically consequential autonomous actions with explicit safety, efficacy, monitoring, and regulatory requirements.

Decision implication

Create a dedicated agentic-AI adoption standard covering authority boundaries, human escalation, post-deployment monitoring, evidence, liability, workflow, and regulatory status.

Who or what is affected

ARPA-H
FDA
cardiovascular health systems
AI developers
clinicians
patients

Evidence and sources

Decision priority: Watch

Source authority: Primary/authoritative · official ARPA-H source

Evidence design: Program record

Applicability: Direct

Confidence for the stated claim: High

Freshness: Changed

P4L distinguishes the sourced record from interpretation. Program inclusion, participation, funding, proposal status, or a public listing does not by itself establish clinical effectiveness, endorsement, coverage, or scaled adoption unless the cited source specifically establishes it.

For this decision

Researchers · Innovators · Funders/investors · Policymakers · Clinicians

What P4L is watching next

Awardees
supervisory-agent architecture
FDA pathway
clinical validation
deployment sites.

Related proof records

CLM-0010 — ARPA-H describes ADVOCATE as intended to develop clinical AI agents for cardiovascular-care tasks plus supervisory post-deployment monitoring.

Record provenance

Last verified: 2026-09-03
Source date (official launch release): 2026-01-13
Revision state: Version 1.1 recasts ADVOCATE as a prospective program objective, records the closed solicitation and no-results boundary, adds the official launch source, and records the September 3, 2026 re-verification.
Status: Prospective program objective · solicitation closed 2026-04-01 · no results reported
Freshness: Changed
Version: 1.1
Stable ID: SIG-2026-017