ARPA-H commits $62.7M to autonomous AI for heart failure care

Federal funding targets AI-driven heart failure management

The Advanced Research Projects Agency for Health has committed $62.7 million to develop artificial intelligence systems capable of autonomously managing heart failure treatment, including prescribing medications and ordering diagnostic tests.

The program, designated ADVOCATE, represents a significant federal investment in AI-powered clinical decision-making. ARPA-H announced its first round of recipients on Wednesday, distributing funds to health technology companies Atman Health, UpDoc, and Tempus AI, alongside research teams from Stanford University, Duke University, and Kaiser Permanente. The agency may expand the program to include additional teams in future funding rounds.

The initial year commitment totals $33.7 million, with the remaining funds subject to renegotiation based on program performance and milestones.

Why it matters

Heart failure affects 6.7 million Americans, many of whom struggle to access specialized care due to geographic and systemic barriers. Autonomous AI systems authorized by the FDA could fundamentally reshape how chronic disease management works in rural and underserved communities, potentially extending specialist-level care without requiring in-person visits. The program also signals federal willingness to pursue regulatory pathways for AI systems that make clinical decisions traditionally reserved for physicians—a threshold that could accelerate or complicate broader AI adoption in healthcare depending on safety outcomes.

Addressing the specialist access gap

The ADVOCATE program directly targets a documented treatment gap in heart failure care. According to ARPA-H, many patients with the condition don't receive optimal management because they cannot easily reach cardiologists or heart failure specialists. This access problem is particularly acute in rural areas and communities with limited healthcare infrastructure.

The AI systems funded through this initiative are designed to assess symptom severity, make prescribing decisions, and determine when laboratory testing is necessary—functions that currently require physician oversight. The goal is to deploy these capabilities in settings where specialist availability is limited or nonexistent.

Regulatory pathway for autonomous medical AI

A central objective of ADVOCATE is achieving FDA authorization for partially autonomous AI devices in clinical care. This regulatory milestone would mark a significant evolution in how medical AI is deployed, moving beyond decision-support tools that assist physicians toward systems that can execute treatment decisions independently within defined parameters.

The program's structure, with initial funding followed by performance-based continuation, suggests ARPA-H is taking an iterative approach to both technology development and regulatory strategy.

Details on the ADVOCATE program were first reported by STAT News.