Smartwatches still cannot reliably measure blood sugar from the wrist. But Apple and Samsung may not need a breakthrough in sensors to make glucose monitoring part of their next AI health battle.
As continuous glucose monitors (CGMs) move further into mainstream wellness, the next wearable contest may be less about detecting glucose directly and more about explaining it. Apple, Samsung, Oura, Dexcom, Abbott, and other health-tech players are circling the same opportunity: turning data on glucose, sleep, stress, exercise, nutrition, and heart rate into personalized AI guidance.
That shift could make glucose monitoring the sharpest test yet for consumer health AI. The promise is useful: real-time coaching. The risk is that a wellness nudge can start sounding a lot like medical advice.
Smartwatches still have a glucose problem
The dream is simple: glance at your wrist and see your blood sugar. The reality is not.
Despite years of rumors and research, today’s leading smartwatches and smart rings do not offer FDA-cleared, noninvasive blood glucose measurement. In 2024, the US Food and Drug Administration warned consumers not to rely on smartwatches or rings that claim to measure blood glucose without skin penetration, citing the risk of inaccurate readings and potentially dangerous health decisions.
That reality creates a challenge for Apple and Samsung. Glucose is one of the most valuable health metrics they could add to their ecosystems, but it is also one of the most medically sensitive. As TechRepublic recently reported, companies including Apple, Samsung, Garmin, and Oura continue to explore glucose-related health features, even as truly noninvasive glucose sensing remains elusive.
So the near-term path may not be a watch that replaces a continuous glucose monitor. It may be a phone-and-watch ecosystem that connects to CGMs, imports glucose data, and uses AI to explain what the numbers mean.
The AI layer may matter more than the sensor
Continuous glucose monitors already generate streams of data. For people with diabetes, that information can be medically essential. For wellness users, it can reveal how meals, sleep, stress, and workouts affect energy and metabolic patterns.
But raw glucose curves are not always easy to understand. A spike after lunch, a late-night dip, or a sluggish recovery after a poor night of sleep may be useful only if the system can place it in context.
That is where AI becomes the real product.
Instead of simply showing a chart, an AI health assistant could summarize patterns, flag unusual changes, connect glucose swings to meals or workouts, and suggest questions to ask a clinician. The device on the wrist would still matter, but the competitive advantage would move to the software layer that interprets the body’s signals.
Apple and Samsung already have the platform pieces
Apple and Samsung have many of the pieces needed to compete in that direction, even if neither company has announced a consumer smartwatch that can directly measure blood glucose without a CGM.
Apple has the Apple WatchiPhone, Health app, HealthKit, and Apple Intelligence. Samsung has Galaxy Watch, Galaxy Ring, Samsung Health, and Galaxy AI. Samsung has also been pushing toward more AI-driven health interpretation, with recent reports describing a redesigned Samsung Health app built around personalized wellness insights, AI-generated summaries, Vitals, Energy Score, heart health, cardio load, and fitness metrics.
Apple’s current AI-health story is less directly tied to glucose. Its recent Apple Intelligence-connected fitness features, including Workout Buddy, point to a broader interest in using AI to personalize health and activity guidance, but Apple has not announced glucose-specific AI advice for consumers.
That makes glucose data a possible next frontier rather than a confirmed product roadmap. A user’s glucose response to breakfast becomes more meaningful when paired with sleep quality, exercise, heart rate variability, medication logs, and stress patterns. AI can stitch those signals into a story.
That is also why the competition is about more than wearables. It is about who becomes the trusted interpreter of personal health data.
Perhaps the clearest example of that future is already here.
In May 2025, Oura expanded its partnership with Dexcom to bring glucose insights from Dexcom’s Stelo continuous glucose monitor into the Oura experience. As The Verge reportedthe feature combines glucose readings with Oura’s sleep, activity, stress, and recovery metrics, then uses AI-generated summaries to help users interpret metabolic patterns.
There are important limits. The feature relies on Dexcom’s Stelo CGM rather than glucose sensing from the Oura Ring itself, and The Verge reported that it was available in the US at launch, with Stelo sold through Oura for $99.
That model is important because it separates sensing from interpretation. The CGM handles the glucose signal. Oura focuses on the AI-powered experience around it.
In other words, the hardware gathers the data, but the AI creates the value.
Apple and Samsung could follow a similar strategy on a much larger scale. They do not necessarily need to win the first battle by building the perfect noninvasive glucose sensor. They can win by making glucose data understandable, useful, and habit-forming.
The FDA line is the hard part
The challenge is that glucose is not a step count. It is not a sleep score. It is a metric people may use to make decisions about food, exercise, medication, and medical care.
That creates a narrow path for consumer technology companies. An AI assistant can summarize trends, encourage users to consult a physician, or provide educational wellness insights. But the closer it gets to diagnosis, treatment, or medication recommendations, the more regulatory scrutiny it may attract.
This is where Apple and Samsung’s usual consumer-tech playbook may not be enough. Health AI has to be useful without being reckless, personalized without being invasive, and persuasive without overstating certainty.
The companies that win this market will likely be the ones that treat trust as seriously as features.
Glucose could become the test case for AI health
AI health assistants are still in their early stages, and much of the market remains filled with ambitious promises. Glucose data could force the category to mature.
Unlike broad wellness scores, glucose readings are specific, continuous, and tied to real health decisions. That makes them valuable for AI systems, but also harder to handle responsibly.
For Apple and Samsung, the opportunity is clear: build the AI layer that helps people understand what their body is doing throughout the day. For users, the question is whether those insights are accurate, private, and clear about their limits.
The smartwatch may not become a blood sugar monitor anytime soon. But the race to turn glucose data into AI health advice is already underway, and the companies that win it may be the ones that never needed to measure glucose themselves.
Editor’s note: This article originally appeared on our sister publication, eWeek.