Federal Initiatives Aim to Move AI Beyond Simple Diagnostics

The program, called Agentic AI-Enabled Cardiovascular Care Transformation (ADVOCATE), aims to establish a pathway toward FDA-authorized agentic AI solutions that go beyond predictive models and toward autonomous, patient-facing assistance in high-risk clinical care settings like cardiovascular disease. Under the plan, selected innovation teams will develop two AI agents designed to support cardiovascular patients and deliver FDA authorization within approximately three years.

Under the program, AI agents would interact with electronic health records, wearable data and patient inputs to assist with tasks such as medication management, appointment scheduling, symptom monitoring and lifestyle guidance between clinical visits. Federal officials have framed the effort to address clinician shortages, rising chronic disease burden and gaps in access to continuous care, including populations with high chronic disease burden.

Unlike earlier FDA-cleared AI tools that focus on narrow diagnostic functions, these systems would be designed to operate autonomously within defined boundaries, escalating issues to clinicians when thresholds are crossed. The deadline to submit a summary proposal is Feb. 27.

CMS Builds Infrastructure for Patient-Facing AI

While ARPA-H and the FDA focus on approval pathways, the Centers for Medicare and Medicaid Services (CMS) is working to reshape the healthcare infrastructure that would allow clinical AI tools to function at scale, according to the news site Fierce Healthcare. CMS has made health technology modernization a central priority for 2026, with an emphasis on interoperability, patient access to data and the expansion of digital tools that support chronic disease management.

Through its voluntary Health Tech Ecosystem initiative, CMS has encouraged healthcare organizations to adopt standardized data-sharing frameworks such as Fast Healthcare Interoperability Resources and to support app-based access to medical records. Agency officials argue that patient-facing AI applications and clinical agents cannot operate safely or effectively without reliable, real-time data flows across hospitals, outpatient providers and home-based settings. Since mid-2025, hundreds of organizations have joined the CMS pledge, signaling growing alignment around common technical standards.

CMS leaders have also highlighted AI’s potential to reduce administrative friction and improve patient engagement, from simplifying intake and eligibility checks to supporting medication adherence and ongoing monitoring for conditions such as diabetes and obesity.

Although CMS does not regulate medical devices, its influence over Medicare and Medicaid payment policy positions it as a critical gatekeeper for widespread adoption. Any future reimbursement models for AI-enabled care will depend on the interoperability and patient-access foundations CMS is now promoting.

FDA Clears Multi-Condition AI as Deployment Accelerates

That convergence of policy, infrastructure and regulation is already visible in the FDA’s recent clearance of a comprehensive AI triage platform for radiology. The agency authorized Aidoc’s aI triage system, which is capable of detecting 14 acute conditions from CT scans, combining multiple indicators into a single workflow rather than approving each condition separately. The clearance marks one of the most expansive authorizations yet for clinical AI and reflects the FDA’s evolving approach to more complex, platform-style models.

Performance data reviewed during the clearance process showed high sensitivity and specificity across the supported conditions, with a reduction in false alerts compared with earlier single-purpose tools. The system is designed to prioritize scans with suspected acute findings, helping radiologists and emergency physicians identify critical cases more quickly in crowded clinical environments. Developers say the platform has already analyzed millions of patient cases in real-world deployments, providing regulators with a growing evidence base for safety and effectiveness.