AI Robotics in Medicine

Israel’s Medical Association released principles for bringing AI into clinical settings, stressing that tools should assist diagnosis and treatment while leaving judgment and responsibility with clinicians, and clarifying it isn’t patient-specific guidance.

Framed against rapid regulatory momentum—by May 2024 the U.S. FDA had cleared 882 AI-enabled devices, 76% in radiology—the paper notes AI use long predates generative models.

Real-world examples in Israel, from Aidoc flagging urgent CT/X-ray findings for faster radiologist review to Rounds transcribing visits and generating summaries to reduce typing and burnout, illustrate augmentation and safety layers rather than replacement.

Tracking: Medicine Robotics · AI Medicine · AI Healthcare

Geography: United States, European Union, United Kingdom, Canada, Israel, China, Japan, South Korea, Singapore, India, Australia

1. Israeli Medical Association sets guardrails for AI in clinical practice

Israel’s Medical Association has issued a position paper outlining general principles for integrating artificial intelligence into care settings, published this month through its quality and risk-management bodies.

The document is explicit that it is not patient-specific guidance and that AI should assist diagnosis and treatment without supplanting clinicians’ judgment or responsibility.

While generative AI accelerated interest, the paper notes AI was already embedded in imaging interpretation, risk assessment, patient prioritization, and EHR recommendations.

The paper cites rapid regulatory momentum: by May 2024, the U.S. FDA had cleared 882 AI-enabled devices, with radiology accounting for 671 (76%).

In Israel, Sheba has implemented Aidoc to flag urgent findings on CT and X-rays so radiologists can review them sooner, adding a safety layer without replacing interpretation.

The Rounds system targets burnout by recording clinical visits, transcribing conversations, and generating visit notes or summaries to reduce typing.

Key facts:

- Israeli Medical Association issued AI integration principles in June 2026; clinicians retain judgment and responsibility.

- As of May 2024, FDA approved 882 AI-enabled devices; radiology accounts for 671 (76%).

- Sheba Medical Center uses Aidoc to triage urgent CT/X-ray findings and alert radiologists.

Why it matters: Clear guardrails help hospitals adopt AI where it is strongest—triage and documentation—while curbing overreliance on opaque systems. Clinicians gain support for imaging backlogs and note-writing, but retain decision-making authority and accountability.

For vendors, the message favors assistive, workflow-embedded tools over fully autonomous decision-makers.

With hundreds of cleared devices competing for clinical attention, success will hinge on safety evidence, seamless EHR integration, and demonstrable reductions in missed findings and documentation burden.

Watch how Israeli providers operationalize governance, training, and post-deployment monitoring to translate principles into consistent, safe practice.

Jpost