Your Clinic Says It Uses AI. How Can You Check if the Tool Is FDA-Authorized?

If a clinic, imaging center, or hospital tells you it uses artificial intelligence, you do not have to settle for a vague answer. A practical next step is to ask which tool is being used and whether it has a public FDA device record.

That matters because “AI in healthcare” can mean very different things. Some tools help read scans or support a screening step. Others help with scheduling, note-taking, or message triage. FDA device authorization applies to specific medical devices or software functions, not to every AI feature a health system may use.

FDA’s Artificial Intelligence-Enabled Medical Devices page was updated on June 16, 2026. Combined with the public Devices@FDA database, it gives patients and caregivers a way to verify at least part of what a clinic is using. But those tools also have limits, and an FDA listing is not the same as proof that a product will work equally well in every hospital, workflow, or patient group.

Start by asking for the exact name of the tool

If someone says, “We use AI,” ask for details in plain language. A useful question is: What is the name of the device or software, who makes it, and what is it supposed to do?

Try to get as many of these identifiers as possible:

Device or software name
Manufacturer name
Model or version, if available
FDA submission number, if the clinic has it
What the tool is used for, such as reading images, analyzing a signal, or assisting with a screening step

The more specific the identifier, the easier it is to check the public record. A broad statement such as “we use AI in radiology” usually is not enough to confirm what product is involved.

Step 1: Use the FDA’s AI-enabled medical devices page

FDA’s AI-enabled medical devices page is the best place to understand what the agency says it is tracking. FDA says the page is meant to identify AI-enabled medical devices that are authorized for marketing in the United States and to improve transparency for patients and healthcare professionals.

The important word is devices. The page is not a master list of every AI product a clinic or hospital might use. If the organization is talking about an AI tool for operations, documentation, scheduling, or another non-device function, the FDA device materials may not answer your question.

FDA also says the list is not comprehensive. The agency explains that the list is mainly built by identifying AI-related terms in public summaries and device classifications, and that some devices may be added later as public decision documents become available. So if you do not find a product on that page, that alone does not prove the clinic is using something improper. It does mean you should ask for more specific identifying information.

Step 2: Search Devices@FDA

Once you have a device name, manufacturer, or submission number, use Devices@FDA. FDA describes this as a public catalog of cleared and approved medical device information, and it can help patients confirm whether a specific product appears in the agency’s records.

A matching record may help you confirm:

That the device has a public FDA record
That the manufacturer matches what the clinic told you
That the intended use fits the way the clinic described the tool
Whether there are public summaries, labeling, instructions, or other materials you can review

If you have a submission number from the FDA AI-enabled device page, that usually makes the search easier. If not, try the product name and manufacturer first. If you still cannot find a clear match, ask the clinic to write down the exact product name and version. Similar names, software updates, and rebranding can make quick searches confusing.

What the public record can tell you — and what it cannot

A public FDA record is useful, but it is only the start of the conversation.

What it can tell you: whether a specific device appears in FDA materials, which company is tied to it, how it is described publicly, and sometimes what summaries or user information are available.

What it cannot tell you by itself: whether the tool performs equally well in every clinic, every patient population, or your individual case. It also does not tell you how your care team uses the tool in practice, what safeguards are in place when the output is uncertain, or how often a clinician overrides it.

That is why “FDA-authorized” should not be treated as the final answer. A 2026 cohort study in JAMA Network Open examined 903 FDA-authorized AI-enabled medical devices and found that missing publicly available clinical-study information and use-related issues were associated with higher recall hazards. This was an observational study, so it cannot prove that missing public information caused recalls, and a recall does not automatically mean patient harm in every case. But it does support a simple point: marketing authorization and real-world performance are not the same thing.

Questions to ask your provider after you check the record

After you look up the device, the next step is not to judge the technology on your own. It is to ask better questions.

Helpful questions include:

What exactly does this tool do in my care?
Is it making a diagnosis, flagging a finding, helping with screening, or supporting another step?
Who reviews the output before any medical decision is made?
What happens if the AI result is uncertain?
What happens if the AI output conflicts with the clinician’s judgment or with my symptoms?
Has this tool been evaluated in patients like me or in a setting like this one?
Can I have the device name or submission number in my after-visit paperwork?

CDC patient-safety guidance encourages patients and caregivers to speak up, ask questions, and talk with healthcare professionals about concerns. That approach fits AI use too. If a result seems confusing, rushed, or inconsistent with how you feel, it is reasonable to ask for clarification.

Why asking matters now

Patients are likely to run into AI more often, but that does not mean every clinic uses it the same way or trains staff the same way. A June 2026 Reuters report on a clinician survey found that many respondents said AI training at their organizations was inadequate, inconsistent, or unavailable. That does not mean the tools are unsafe by default. It does mean patients may encounter AI in real workflows even when the details are not well explained.

The AMA has also emphasized that transparency, accountability, and meaningful physician oversight matter as AI becomes more common in care. For patients and families, the practical takeaway is simple: if a tool could influence a diagnosis, screening result, or treatment path, ask who is checking it and what backup plan exists if the software is wrong, uncertain, or incomplete.

A short checklist to bring to your appointment

If you want to keep this simple, bring this checklist with you:

Ask for the exact device or software name
Ask for the manufacturer name
Ask whether there is an FDA submission number or public device record
Look for the device on FDA’s AI-enabled medical devices page or in Devices@FDA
Ask what the tool is used for in your care
Ask who reviews the output before action is taken
Ask what happens if the result is uncertain or conflicts with clinical findings
Write down the answer or ask for it in your visit summary

If your symptoms are worsening or feel urgent, do not wait for a software result alone. Contact your care team promptly. Use emergency care right away for serious warning signs such as chest pain, severe trouble breathing, stroke symptoms, severe bleeding, or other emergency symptoms.

The bottom line: when a clinic says it uses AI, the safest response is neither panic nor blind trust. It is specific verification, clear questions, and an understanding that human oversight still matters.

Sources

Editorial note: Weence articles are researched from cited public-health, medical, regulatory, journal, and reputable news sources and may be drafted with AI assistance. They are checked for source support, clarity, and safety guardrails before publication.

This article is for general informational purposes only and is not medical advice. Research findings can be early or incomplete, and health guidance can change. Always talk with a qualified healthcare professional about personal symptoms, diagnosis, medications, vaccines, screenings, or treatment decisions. If you think you may have a medical emergency, call emergency services right away.