Photo by freestocks on Unsplash
The FDA's Software as a Medical Device (SaMD) line comes down to one question: does the app diagnose, treat, or prevent disease — or does it just track and encourage? A Yonkers-based startup building a symptom tracker for perimenopause ran into that exact question this month when its product team wanted to add a "risk score" feature. As of July 22, 2026, according to Google News, outlets including Yonkers Times have been covering a wave of similar cases as digital health developers try to figure out which side of the FDA's SaMD line their app actually sits on.
What's on the Table
Software as a Medical Device, or SaMD, is the FDA's term for software that serves a medical purpose independent of any hardware device — no sensor, no wearable, just code making a clinical judgment. The definition traces back to guidance from the International Medical Device Regulators Forum (IMDRF), adopted by the FDA in 2013-2014. The test isn't the technology; it's the intended use. Software that diagnoses, treats, cures, mitigates, or prevents disease falls under FDA oversight, governed in part by 21 CFR Part 820, the agency's quality system regulation (the rules covering how medical device manufacturers must design and document their products).
The FDA's 2022 Digital Health Policy Navigator drew a clearer boundary for everyday consumers: general wellness apps, fitness trackers, and health education tools are generally not treated as medical devices. Separately, the FDA's Mobile Medical Applications guidance explicitly exempts categories like appointment schedulers and health record platforms that don't offer decision support. Clinical decision support (CDS) software is a different story — as of 2024-2025, tools that hand treatment recommendations to healthcare providers typically need 510(k) clearance (a premarket review showing a device is comparable to one already approved) or De Novo classification before they can go to market.
Side-by-Side: How They Differ
The gap between "wellness app" and "regulated medical device" is wider than most founders expect — until it isn't. The FDA's enforcement discretion policy exempts an estimated 80-90% of consumer health apps from medical device regulation entirely. That leaves a meaningful slice of the market that does need to clear regulatory review, and the numbers there are substantial: the FDA's Digital Health Center of Excellence reported 801 total digital health device marketing submissions between 2011 and 2023, with 90% of them receiving 510(k) clearance. More than 600 digital health and SaMD products received FDA clearance or approval between 2015 and 2024 alone.
Within that regulated slice, risk tiering matters. FDA guidance from 2023 splits software into three classes: Class I (lowest risk, sometimes exempt from premarket review), Class II (requires 510(k) premarket notification), and Class III (requires full premarket approval, or PMA). Class II software — think ECG-analysis apps and diabetes management tools — makes up roughly 75% of FDA-cleared SaMD submissions, making it the workhorse category for digital health.
Notably, the FDA's own guidance documents are more precise on this than most secondhand coverage of the topic; Yonkers Times' original item on the 2026 SaMD line wasn't fully accessible for this piece, underscoring how much of the practical clarity here still comes straight from the regulator rather than from news summaries. The FDA tried to shortcut this process once before: its Pre-Cert pilot program, launched in 2017, aimed to fast-track approval for established digital health companies. It ended in 2022 without becoming permanent policy. Meanwhile, the European Union's Medical Device Regulation (MDR) and In Vitro Diagnostic Regulation (IVDR) phased in stricter requirements between 2021 and 2024, adding a second, non-matching compliance track for any developer selling on both continents.
The AI Angle
Artificial intelligence is where the SaMD line gets genuinely hard to see. An algorithm that passively logs your steps or sleep is a wellness tool. The same algorithm, retrained to flag atrial fibrillation or predict diabetic retinopathy, crosses straight into medical device territory — a distinction the FDA addressed directly in its AI/ML-based SaMD action plan guidance, finalized in 2023-2024. That guidance introduces oversight for "continuous learning" algorithms and requires developers to file predetermined change control plans spelling out exactly how a model is allowed to evolve after it's cleared. AI/ML-enabled devices have grown roughly 40% annually since 2020, and that pace is part of why regulators felt the need to write rules for models that keep changing after approval, rather than shipping once and staying static.
For investors tracking digital health as part of a broader investment portfolio, this is worth watching closely: AI investing tools that screen healthcare stocks are increasingly factoring in SaMD regulatory exposure — whether a company's flagship product sits in the 80-90% exempt bucket or the smaller, costlier regulated tier — alongside more traditional metrics like revenue growth.
Which Fits Your Situation
Digital health lawyers who work this terrain tend to agree on the tiebreaker: marketing claims, not code. As one puts it, "The line between wellness and medical device hinges on marketing claims—if you claim to diagnose or treat, you're likely a medical device regardless of technology." Industry consultants echo the risk from the founder's side, noting that "many startups unknowingly cross the SaMD threshold by adding clinical decision features to wellness apps, triggering unexpected FDA jurisdiction."
That has real financial planning consequences. A wellness app that adds a single "your risk of X is elevated" feature can accidentally trigger 510(k) review timelines that run months and cost real money — a detail that belongs in any founder's financial planning, not just their product roadmap. For everyday users, the practical takeaway is smaller but still useful: an app's FDA status is a rough proxy for how rigorously its health claims have been vetted, which matters if you're folding an app's output into decisions about your own personal finance around healthcare spending, insurance, or provider visits.
On balance, our analysis suggests the SaMD line will keep getting blurrier before it gets clearer — not because the FDA's framework is broken, but because AI models are evolving faster than any static classification system was built to handle. The more likely outcome is incremental guidance updates, not a single bright-line rewrite, so developers and investors alike should expect to keep reading the fine print app by app, feature by feature.
Frequently Asked Questions
Is a symptom-tracking app considered a medical device by the FDA?
It depends on what the app claims to do. If it only logs symptoms for the user's own reference, it typically falls under the FDA's enforcement discretion as a wellness tool. If it generates a diagnosis, risk score, or treatment recommendation, it likely qualifies as SaMD and may need 510(k) clearance or De Novo classification.
Do wellness apps need FDA approval to be sold in 2026?
Most do not. The FDA's enforcement discretion policy covers an estimated 80-90% of consumer health apps, exempting general fitness, wellness, and health education software from medical device regulation.
What's the difference between Class I, Class II, and Class III medical device software?
Per FDA guidance, Class I is the lowest-risk tier and may qualify for exemptions; Class II requires 510(k) premarket notification and covers roughly 75% of cleared SaMD submissions, including tools like ECG-analysis apps; Class III, the highest-risk tier, requires full premarket approval (PMA).
Disclaimer: This article is for informational purposes only and does not constitute financial or medical advice. Research based on publicly available sources current as of July 22, 2026.