A lot of health apps sound medical. Fewer are actually regulated like medical devices.

That distinction matters more now that AI features are showing up in symptom checkers, wellness trackers, coaching apps, and doctor-facing software. The FDA has been pretty clear that not every health-related app becomes a “mobile medical app” just because it talks about health. [1]

The simple version: FDA cares about what the software does

The FDA focuses on “software functions,” not just whether an app lives on your phone. [1]

An app usually starts attracting FDA oversight when it’s intended to diagnose, treat, cure, mitigate, or prevent disease, or when it turns a phone into part of a medical device. [1]

Examples the FDA has historically treated more like medical-device software include apps that:

  • analyze medical images,
  • calculate medication dosing,
  • connect to medical sensors,
  • monitor patient conditions in ways clinicians rely on,
  • or provide treatment recommendations. [1][3]

Meanwhile, a lot of everyday health software falls outside stricter FDA regulation:

  • fitness trackers,
  • habit logging,
  • wellness journaling,
  • appointment scheduling,
  • general education,
  • or apps that just organize health information. [1]

That’s why two apps can both say they “help improve health,” but only one gets treated like regulated medical technology.

AI health tools are making the line fuzzier

This is where things get messy for consumers.

A chatbot that gives generic wellness advice is different from software making patient-specific treatment recommendations. The FDA and health-policy groups keep pointing back to the same core question: can a clinician or user independently review the basis for the recommendation, or is the software effectively acting like a medical decision-maker? [3][4]

Clinical Decision Support (CDS) software is a good example. Some CDS tools can avoid being regulated as medical devices if they meet certain legal criteria, including allowing healthcare professionals to independently evaluate the recommendations. [4]

But once software becomes more opaque, automated, or central to diagnosis and treatment decisions, FDA scrutiny tends to increase. [3][4]

That’s also why you’re seeing the term “Software as a Medical Device” (SaMD) appear more often in healthcare discussions. CMS even references FDA oversight of SaMD and mobile medical apps in its own technical guidance. [2]

App stores are not the same thing as FDA review

This part trips people up constantly.

Apple reviewing an app for the App Store does not mean the FDA evaluated its medical claims. Apple’s review process focuses heavily on safety, security, privacy, malware prevention, and platform rules. [6]

That’s useful, but it’s a completely different system from FDA medical-device oversight.

So if an app says it uses AI to detect a health condition, consumers should separate:

  • “approved in the App Store”
    from
  • “reviewed or regulated by the FDA.”

Those are not interchangeable.

The bigger shift: FDA is treating digital health as its own category now

The FDA’s Digital Health Center of Excellence exists partly because health software no longer fits neatly into old medical-device categories. [5]

Phones, wearables, cloud software, and AI tools now blur together. A smartwatch can collect medical-grade signals. A tablet app can assist with clinical workflows. A chatbot can influence care decisions.

The practical takeaway for normal users is pretty simple:

  • not every health app is a medical device,
  • medical-sounding marketing does not automatically mean FDA-reviewed,
  • and the more software directly influences diagnosis or treatment, the more likely FDA rules start applying. [1][3][5]

That’s usually the better signal to watch than whether an app calls itself “AI-powered.”

Sources