MedicalRecords Research · Source guide
Abridge vs Suki AI: research checklist
A source-backed framework for comparing healthcare AI tools, with questions about evidence, intended use, and implementation.
Source review: · October 2026 source review
What the sources establish
Source note 1 · official data
The FDA publishes a list of AI-enabled medical devices authorized for marketing in the United States, with links to device records. The FDA says the list is not comprehensive. It is not a census of all healthcare AI software or a market-share dataset.
- FDA list of AI-enabled medical devices — U.S. Food and Drug Administration
Scope and evidence limits
General source guidance, not a finding about a named vendor, a patient, or an individual records request.
This edition publishes only the sourced framework below. Earlier numerical estimates, product rankings, comparative outcomes, and reviewer endorsements are withheld pending claim-level verification. Source review does not establish current vendor pricing, individual eligibility, legal outcomes, clinical suitability, or service availability.
Confidence: high. The statement is limited to the contents and purpose explicitly described by the cited primary source.
Source note 2 · official data
NIST provides a voluntary AI Risk Management Framework to help organizations address risks associated with AI. The framework supports risk-management processes; it does not certify or rank the healthcare vendors in this directory.
- AI Risk Management Framework — National Institute of Standards and Technology
Scope and evidence limits
General source guidance, not a finding about a named vendor, a patient, or an individual records request.
This edition publishes only the sourced framework below. Earlier numerical estimates, product rankings, comparative outcomes, and reviewer endorsements are withheld pending claim-level verification. Source review does not establish current vendor pricing, individual eligibility, legal outcomes, clinical suitability, or service availability.
Confidence: high. The statement is limited to the contents and purpose explicitly described by the cited primary source.
Questions for your comparison
An editorial planning checklist. These questions are not product findings, medical recommendations, or promises of service.
- What exact use case and product version are you evaluating?
- Which claims are vendor statements, and which have independent evidence?
- What data is retained, who can access it, and what terms cover your organization?
- What would a documented pilot need to demonstrate before wider use?
Continue your research
Publication scope
This edition publishes only the sourced framework below. Earlier numerical estimates, product rankings, comparative outcomes, and reviewer endorsements are withheld pending claim-level verification. Source review does not establish current vendor pricing, individual eligibility, legal outcomes, clinical suitability, or service availability.
Read the linked primary sources on October 11, 2026; publish only narrow statements about what those sources provide. The checklist is editorial planning guidance, not a measured vendor or treatment evaluation.
Review record
Codex — automated primary-source review; no human editorial, legal, clinical, or statistical sign-off asserted
Version 2.0.0. 2 source records; 2 statements.