FDA’s General Wellness Shift: De-Risking AI for Investors

The world of digital health regulation is anything but static, especially where it smashes into the fast-moving front of AI. If you’re on a product or compliance team trying to build in this space, you have to stay on top of regulatory shifts. A great example of this is how the FDA recently re-thought its general wellness policy, a process kicked off by a single enforcement action and resulting in new guidance that gives a much clearer lane for non-invasive physiological signal estimation.

The Warning and its Immediate Aftermath

In July 2025, the FDA sent a warning letter to WHOOP about its Blood Pressure Insights (BPI) feature. The letter wasn’t entirely public, but the message was clear: the agency felt the BPI feature’s claims might be stepping into medical device territory. Back then, the common wisdom was that any product claiming to measure or estimate a physiological signal like blood pressure, even just for “wellness,” was at high risk of being classified as a medical device, which would mean getting tangled up in premarket review (like a 510(k) or De Novo) and needing a full QMS / ISO 13485-compliant system. That letter sent a shockwave through the industry. Suddenly, companies with AI features that gave users health insights but stopped short of actual diagnosis were in a tough spot. The line between a “general wellness” product and a “medical device” got very blurry, and nobody knew what the right regulatory path was for these new non-invasive monitoring tools. The big fear was that a super-broad interpretation from the FDA could kill innovation in wellness tech, forcing a bunch of products meant for general health awareness into the expensive and time-consuming medical device pipeline.

The Guidance: A Policy Pivot

Then, less than six months later in January 2026, the FDA did an about-face and released updated general wellness and Clinical Decision Support (CDS) guidance. This was a real policy shift, specifically written to cover the kinds of features that had just gotten them in trouble. The new guidance made it explicit: non-invasive products that estimate physiological signals, including blood pressure, can be considered general wellness devices, but only if they follow some very specific rules. MedTech Dive report on FDA guidance January 2026 For any product or compliance team, you absolutely have to know the agency’s criteria for making a blood pressure-adjacent wellness claim under this new policy:

• Intended Only for Wellness: The product’s intended use has to be for general wellness, period. Think promoting a healthy lifestyle or just making someone more aware of their body’s patterns. It can’t be for diagnosing, treating, mitigating, or preventing any disease or condition.
• Non-Invasive and Low-Risk: The device can’t break the skin and has to be low-risk for the user. This is a foundational piece of the general wellness classification that separates these products from more serious medical devices.
• Avoiding Disease, Diagnostic, or Clinical-Management Claims: The product must stay far away from any claims about diagnosing or treating a disease, or managing a clinical condition. This is the most important distinction. For example, a pop-up saying “Your estimated blood pressure is elevated. Consult your doctor for hypertension diagnosis” would almost certainly be a regulated medical device claim. One that says “Your estimated blood pressure is in a higher range today, consider stress reduction techniques” might just stay on the wellness side of the line.

This update finally provided some much-needed clarity. It created a real pathway for products that give people useful physiological insights without pretending to be a doctor. The FDA acknowledged that there’s value in giving people more data about their own bodies, as long as the tech is sold and presented responsibly.

The Nuance of “Responsible AI” in Wellness

The FDA’s new position really draws a line between just letting any AI run wild and using clinically validated, responsibly built AI in the wellness world. The initial warning to WHOOP was about the risk of a feature being misinterpreted as medical advice, but the guidance that followed created a playbook for how to use AI the right way within that general wellness box. In fact, the FDA has now formally dropped its enforcement action against WHOOP’s BPI, stating that, consistent with the January 2026 guidance, it doesn’t plan to enforce device requirements against the feature. This shows the agency is actually paying attention to new tech and is willing to adjust its own rules to let innovation happen without compromising public health. It all comes down to the intent and the claims you make. An AI model that estimates blood pressure to help a user track trends for their own awareness, without telling them what to do about it clinically, fits the new policy. On the other hand, an AI chatbot that gives out incorrect drug interaction advice or an AI system that causes a user to undertriage a cardiac emergency will always be a high-risk medical device that needs serious regulatory oversight, like a 510(k) or even a De Novo. The agency is learning that not all AI-powered health insights are the same. The rules are changing to create a difference between AI that helps with general health awareness and AI that’s doing the job of a medical device or a clinician. This is a critical distinction for anyone building these products, you have to design your AI’s features, its UI, and all of your marketing copy to match the regulatory category you’re aiming for.

Implications for Product and Compliance Teams

If you’re working on or even just thinking about a BP-adjacent product, this whole saga is a stark reminder of how fast the regulatory ground can move. The fact that the FDA went from a formal warning letter to a complete policy revision in about six months shows that they’re trying to keep up with the market. Arnold & Porter analysis of FDA guidance So what should you and your teams be doing right now?

• Vigilance on Regulatory Updates: FDA guidance documents aren’t static. They’re living things. You have to constantly monitor for updates, especially in fast-moving fields like AI and digital health. This isn’t optional.
• Precise Claim Scrutiny: Every single word you use to describe your product, from your website copy to the little pop-ups in your app, has to be scrubbed against the current guidance. The line between a “wellness insight” and a “diagnostic aid” is thin, and crossing it has huge regulatory consequences.
• Intent and Risk Assessment: Be brutally honest about the intended use of your AI features. Is this really for general wellness, or does it start to look like diagnosis, treatment, or clinical management? You need a thorough risk assessment that thinks through how a user could misuse or misinterpret the information you’re giving them.
• Design for Compliance: You have to build your AI products with the regulatory path in mind from day one. That could mean preparing for QMS / ISO 13485 if it’s likely to be a medical device, or it could mean putting up strict guardrails to ensure your “general wellness” product stays in its lane.

This whole episode shows the FDA trying to get out in front of AI in health, trying to find a way to let good ideas flourish without putting people at risk. The agency is actively shaping the rules for new technologies, not just waiting for things to go wrong. The story of the FDA’s changing view on blood pressure estimation is a perfect illustration of how policy evolves. It shows how an enforcement action can clarify the rules we have now, but also kick off a bigger conversation that leads to better, more practical guidance down the road. FDA general wellness policy reissued Jan 2026 For product and compliance teams, the lesson here isn’t just about reading the letter of the law. It’s about understanding the dynamic nature of regulatory oversight in the age of AI.