Just because smartwatches can spot dementia risk, should they?

If CES has shown us anything, it's that wearables are getting more sophisticated by the day. Thanks to the advancement of AI/LLMs, we now have pendants that track UV light exposure and rings that listen to every word that's said around us and transcribe it to us.

One company has been at the forefront of the wearable health revolution for years: Samsung. The South Korean company has a clear vision for its tech, which it hopes will watch over your long-term health, spotting subtle changes before you do.

And while Samsung's smartwatches are really good at monitoring cardiovascular and even metabolic health, one area they haven't touched yet – until now. As the brand revealed at CES, cognition and general brain health are the next frontiers, including the detection of early signs of dementia.

On paper, it sounds responsible. Almost as futuristic as the camera that watches you eat. If a device can pick up tiny shifts in speech, movement or daily patterns long before symptoms feel noticeable, isn’t that a good thing? Maybe. But it also forces us to confront a harder question: just because we can measure something this heavy, should we?

The promise is seductive

Early detection has genuine upsides. Doctors already know that spotting cognitive decline sooner can help people access support earlier, start treatments that may slow deterioration, and put plans in place while there’s still time to make choices.

It's also true that you and your loved ones – whether it's friends or family – can get clarity faster. Research gets better data. Since wearables are already trusted companions, quietly collecting context in the background, it makes sense to get them to track more than just steps and heart rate.

From a purely tech perspective, extending that capability to brain health feels like a natural next step. Heart health alerts normalised the idea that your watch might save your life. Why shouldn’t the same apply to your mind?

The cost of knowing too early

Unlike a high heart rate and potential sleep apnea, a notification about “possible cognitive risk” isn’t something you shake off. False positives are inevitable, given that algorithms are not diagnoses.

Yet the moment a device whispers that something might be wrong with your brain, that you might show early signs of dementia, it changes how you see yourself.

Every misplaced key suddenly feels like confirmation. A step missed on the way up to your flat feels like another clue to your slow and inevitable mental decline.

Will Samsung provide you with support when it detects dementia? It's more likely you'll get a standard wearable response, namely to contact your local health professional. And the healthcare pathways that should catch, explain and support people are often slow, inconsistent, and overstretched.

There’s also a bigger ethical shadow here: are we risking medicalising perfectly normal ageing? Forgetfulness happens. Fatigue happens. Life happens. If every natural wobble is reframed as pre-dementia, you don’t get reassurance: you get a lifetime of low-grade dread.

My Oura Ring 4 times says I'm stressed because my respiration and body temperature are higher. It doesn't know that I'm sitting in a warm room in a jumper, or on a crowded bus with the heating on (and my puffer jacket).

Wearables don't see your environment; they can only guess what's happening based on sensor data. Applying it to dementia symptoms, no matter how advanced that machine learning algorithm is, feels risky, and maybe even unnecessary.

Data that cuts very, very deep

This isn’t just sleep stages or VO₂ max. Brain-health signals are among the most intimate data a company could hold. Who controls it? Will insurers want access? Employers? Could a vague “risk score” follow you around the way credit ratings do?

Tech companies love to say 'your data is anonymous,' but history shows that anonymity rarely lasts for long, and we're already giving away so much health data, essentially for free.

Information on your brain health isn’t something you can easily change if it leaks, is misused, or misunderstood. Given that the tech industry has a habit of shipping features first and figuring out the consequences later, a feature that monitors dementia symptoms without any precautions feels borderline reckless.

If wearables are going to flag something this serious, they need to come with human guardrails: clear opt-ins, even clearer explanations, access to qualified clinicians, mental-health support, and realistic language. Not marketing copy or AI reassurance; real pathways to help and support systems.

Choice matters more than capability

None of this means brain-health tracking is inherently bad. For some people, knowledge is power. They’ll want every metric and every early signal. That’s valid.

But equally valid are the people who say: not yet. Who would rather live well now than carry a diagnosis-shaped cloud before anything has actually changed?

That’s where the real question lives. Not “can a wearable detect it,” but “do I want to know, and what happens next if I do?”

Until tech companies can answer that honestly, with strong privacy protections, medical clarity, and genuine support, brain-health wearables sit in an uneasy space: impressive, potentially useful, and deeply, uncomfortably premature.

To be honest, that applies to most of the new applications arriving with AI and large language models. In health tech, the bravest stance isn’t always building the feature first. Sometimes, it’s stepping back and asking whether measuring it is the right thing to do at all.