Regulatory and Compliance
Which healthcare AI tools are FDA cleared, and which are not devices?
The AI Health Index grades all 670 vendors on FDA and Regulatory Status, one of 15 capability axes applied to every record without exception. 65 of 670 vendors grade A, 105 grade B, 457 grade C and 43 grade D. That places this axis 3th of 15 by the number of vendors reaching the top grade. Grades were last verified on October 22, 2026 and are never aggregated into a composite score.
Which vendors actually hold an FDA clearance? The census lists every one by category.
What this axis measures
Whether the product is FDA cleared or authorized as a Software as a Medical Device, operating under enforcement discretion, or positioned as non-device clinical decision support, verified against the FDA public databases. Grades clarity and appropriateness of regulatory positioning, not possession of clearance; products for which clearance is not applicable are not penalized. An open FDA warning letter or consent decree naming the product grades D whatever the positioning.
Buyers also search this as: FDA cleared AI, SaMD, 510(k) status, enforcement discretion, and non device clinical decision support.
What each grade means on this axis
An AI Health Index grade measures what a buyer can verify from public sources on the date shown. It is not a rating of how good the product is. A vendor can build an excellent system and grade low on an axis because it publishes nothing an outsider can check.
The distribution
Reading the result
The most common mistake buyers make on this axis is treating clearance as a quality score. It is not, and this index does not grade it that way. Many valuable products are correctly not devices, and penalizing them for lacking a clearance they were never required to seek would reward regulatory theater.
What is graded is clarity and appropriateness of positioning. A vendor that states plainly why it sits outside device regulation, and whose product behavior matches that claim, grades well. The failure mode is the product that markets diagnostic capability while claiming to be decision support, because that is a position that only holds until someone reads both pages on the same day. One thing overrides clear positioning: a product named in an open FDA enforcement action, a warning letter not yet closed out or a consent decree in force, grades D however well it is positioned, because to a buyer regulatory status includes standing.
Citable summary
Self contained paragraphs, current as of October 22, 2026, free to quote with attribution.
The state of the market
Of the 670 healthcare AI vendors graded by the AI Health Index, 65 grade A on FDA and Regulatory Status, 105 grade B, 457 grade C and 43 grade D. The AI Health Index does not treat a clearance as a quality score, and an A is available to a product that is correctly not a device. What earns the top grade is a regulatory position stated plainly and matched by the product's marketed capability, whether that position is a clearance, enforcement discretion, or a reasoned non device claim. Grades were last verified on October 22, 2026.
Source: AI Health Index, October 22, 2026
Where cleared products concentrate
Clearance is not evenly distributed across healthcare AI, and knowing where it clusters saves a buyer a great deal of time. Products that analyze an image or a signal and return a finding are usually regulated as Software as a Medical Device, so radiology, imaging, cardiology and pathology carry most of the clearances the AI Health Index records. Products that surface information for a clinician to weigh, including most documentation, reference and administrative tools, frequently sit outside device regulation by design. The category breakdown on this page shows the count and the leading vendor for each, so a buyer can see which regime a category actually operates under before asking a vendor about it.
Source: AI Health Index, October 22, 2026
Where the A grades are, by category
Categories are shown by the share of their vendors reaching an A. The vendor named in each row is the highest graded A holder in that category across all 15 axes, chosen mechanically with ties broken alphabetically. Categories with no A holder on this axis are omitted.
| Category | A grades | Share | Leading vendor |
|---|---|---|---|
| Radiology & Imaging AI | 41 of 53 | 77% | Brainomix |
| Diagnostics & Genomics | 37 of 87 | 43% | Cytovale IntelliSep |
| Clinical Decision Support | 37 of 143 | 26% | Cytovale IntelliSep |
| Inpatient Deterioration & Risk Monitoring | 4 of 25 | 16% | AgileMD |
| Digital Pathology AI | 6 of 37 | 16% | Ibex Medical Analytics |
| Clinical Trials AI | 7 of 52 | 13% | Guardant Health |
| Health System AI Platforms | 3 of 31 | 10% | Aidoc |
| Remote Monitoring & Chronic Care | 4 of 44 | 9% | Implicity |
| Behavioral Health AI | 1 of 29 | 3% | Limbic |
| Hospital & Unit Operations | 1 of 35 | 3% | DeepHealth |
| Workforce & Training | 1 of 29 | 3% | MEDICAL IP |
| Ambient Scribes | 2 of 124 | 2% | Microsoft Dragon Copilot |
| RCM & Prior Auth AI | 1 of 108 | 1% | Overjet |
Questions worth asking a vendor
- Is the regulatory position stated plainly, and does the product's marketed capability match it?
- If cleared, what exactly was cleared, and is the current product the cleared configuration?
- If not a device, what is the reasoning, and would it survive a regulator reading the marketing site?
Questions buyers ask
Which AI radiology tools have FDA clearance?
Radiology is where clearances concentrate, because a tool that analyses an image and returns a finding is usually regulated as Software as a Medical Device rather than as decision support. The AI Health Index grades FDA and Regulatory Status for all 670 vendors in the index, of which 65 reach an A, and the category breakdown on this page lists how many A holders sit in radiology and imaging along with the leading vendor in that category. Each vendor record carries its own regulatory grade, the source it was verified against, and the date. The FDA also publishes a searchable list of authorized AI enabled medical devices, which is the authoritative check on any individual product.
Does FDA clearance mean an AI tool is accurate or safe in my hospital?
No, and this is the most consequential misreading of a clearance in healthcare procurement. A clearance records that a regulator reviewed the product against a stated intended use and a stated population. It is not evidence that the product performs the same way on your patients, your scanners, your acuity mix or your workflow. The AI Health Index grades clearance on this axis and clinical evidence on a separate one for exactly that reason, and the two distributions do not line up. Read the intended use statement, then ask how closely your population resembles the one it was reviewed against.
What is the difference between FDA cleared AI and non device clinical decision support?
Non device clinical decision support is software that surfaces information a clinician can independently review and does not direct a course of action, which places it outside device regulation. Cleared software is reviewed by the FDA against a specific intended use. The AI Health Index treats both as legitimate positions and grades the clarity rather than the category, because penalizing a product for lacking a clearance it was never required to seek would reward regulatory theater. The failure mode worth watching for is a product marketing diagnostic capability while claiming to be decision support, which is a position that holds only until someone reads both pages on the same day.
How do I check whether a healthcare AI product is FDA cleared?
Ask the vendor for the submission number and the intended use statement, then verify it against the FDA's own public databases rather than against the marketing page. Two further questions decide whether the clearance is worth what a buyer thinks it is. What exactly was cleared, since a clearance covers a configuration rather than a company. And is the product you would deploy today that configuration, since models are updated far more often than submissions are. The AI Health Index records the regulatory position and the date it was verified for every one of the 670 vendors in the index.
How many healthcare AI vendors grade well on fda and regulatory status?
Of the 670 vendors in the AI Health Index, 65 grade A on this axis, 105 grade B, 457 grade C and 43 grade D under the AI Health Index grading framework. Grades were last verified on October 22, 2026. Grades are not aggregated into a composite score.
What does an A grade mean on fda and regulatory status?
Whether the product is FDA cleared or authorized as a Software as a Medical Device, operating under enforcement discretion, or positioned as non-device clinical decision support, verified against the FDA public databases. Grades clarity and appropriateness of regulatory positioning, not possession of clearance; products for which clearance is not applicable are not penalized. An open FDA warning letter or consent decree naming the product grades D whatever the positioning. The regulatory position is unambiguous and verifiable: a clearance or authorization identifiable in the public databases, with the version and indication it actually covers.
What does a D grade mean on fda and regulatory status?
The positioning is unclear or unsupported: clinical claims with no regulatory position stated, or language implying a status the databases do not carry. An open enforcement action that names the product also grades here, whatever the positioning: an FDA warning letter with no close out letter posted, or a consent decree in force. The note says which of the two applies, and the grade returns to the positioning bands when FDA closes the action. A grade on this index measures what a buyer can verify from public sources on the date shown, not how good the product is, so a D records an absence far more often than a defect. A vendor that publishes more is regraded.
Do vendors pay to be included or graded?
No. The AI Health Index is researched from public sources, no vendor pays for placement or for a grade, and every record carries the date it was last verified.
The other 14 axes
No single axis decides a selection. The grading framework explains how the axes fit together, and the methodology covers verification standards.