Summary
- The technologies enabling autonomous clinical AI are advancing faster than the policy, payment, evidentiary frameworks, and organizational readiness, needed to support their adoption.
- Participants identified meaningful pathways forward but also surfaced unresolved tensions that will require new thinking as autonomous AI capabilities expand across clinical domains.
Key Takeaways
In November 2025, PHTI convened senior leaders from health systems, health plans, technology developers, academia, investment firms, and federal agencies—including clinical experts—for a workshop to explore what is needed to scale AI for autonomous healthcare delivery.
The workshop discussion spanned a common set of questions across two use cases—autonomous hypertension management and mental health chatbots:
Adoption and Evaluation of Progress
- What would need to be true for clinicians to accept and adopt these tools?
- How will the market know whether these tools are driving meaningful progress?
- What factors would encourage payers to cover clinical-grade solutions?
- What would help patients feel confident and safe using them?
Market and Policy Enablers
- What barriers would limit effective adoption today?
- How might stakeholders work to accelerate adoption?
- What insights from these case studies inform regulation more broadly?
Key Themes
The opportunity for AI in hypertension and mental health care
The workshop focused on the requirements for safe, effective, and scalable use of clinical AI—with autonomous prescribing for hypertension management and mental health chatbots as illustrative use cases.
Evidence standards should compare AI to current standards of care and scale with clinical risk.
Evidence standards should compare AI to current standards of care and scale with clinical risk.
Evidence requirements must be rigorous enough to build trust, yet practical enough to avoid stalling innovation. This means having different evidence standards based on the risk of using the AI tool. Autonomous AI tools should be compared to local conditions and the care that patients receive today, not to idealized care. For many, the alternative may be poor access or no treatment at all.
Performance benchmarks should be based on clinical outcomes, and safety standards should adapt as the evidence grows.
Performance benchmarks should be based on clinical outcomes, and safety standards should adapt as the evidence grows.
Ambiguity about what constitutes “good” performance remains a barrier to adoption. Metrics must be anchored to specific, measurable, and meaningful clinical outcomes, rather than to process measures.
New technologies may be initially tested in lower-risk populations but should scale quickly to high-risk populations to maximize impact.
New technologies may be initially tested in lower-risk populations but should scale quickly to high-risk populations to maximize impact.
Lower-risk patients offer tempting on-ramps, but the greatest opportunities for clinical benefits from AI-enabled solutions come from reaching the highest-need patients, including those with higher-complexity conditions and in underserved communities. Reaching these populations may require higher evidence expectations and carry more significant clinical risk.
Widespread adoption will depend on building clinician confidence, gaining clarity about legal liability, and aligning payment models.
Widespread adoption will depend on building clinician confidence, gaining clarity about legal liability, and aligning payment models.
Even highly effective clinical AI faces adoption resistance in the form of norms and culture, concerns about liability, and misaligned incentives. Many health systems are uneasy moving from “some human involvement” to “little-to-no human involvement.” These tensions are impacting regulatory frameworks and the evolution of clinicians’ roles as adoption grows.
Clinical AI: Evidence and Policy Requirements for Scaling Adoption
AI Applications in Healthcare
PHTI convened a three-part workshop series in Washington, D.C., bringing together senior leaders from health systems, health plans, technology developers, academia, investment firms, and federal agencies to explore pathways for responsible AI adoption across healthcare.