Understand · Age Watchers article

AI Is Coming for Healthy Ageing. Older People Deserve a Say.

AI could help us monitor health, stay independent, spot risks and make everyday life easier. But healthy ageing should not mean surrendering control to an algorithm.

Imagine your home ten years from now.

Your watch notices that your walking pace has changed. Your bathroom detects an unusual weight trend. A home sensor notices that you have not moved around normally this morning. An assistant reminds you about medication. Your GP's computer identifies a pattern in your health record. Your car notices your attention drifting. Your phone detects something unusual about your speech.

Some of these technologies already exist in various forms. Others are developing quickly.

Used well, they could help people stay healthier, identify problems earlier, remain independent, manage conditions, stay connected and navigate an increasingly complex health system.

But there is another question:

Who decides what gets monitored — and what happens to the information?

That is where healthy ageing and artificial intelligence collide.

MEASURE. UNDERSTAND. IMPROVE. MAINTAIN.

AI should not decide what healthy ageing looks like

The usual question is often: Can older people keep up with AI?

That question starts with an assumption. Older adults are not one technological ability group.

A 70-year-old may build software, run a business or use ChatGPT every day after four decades of working with computers. Another person of exactly the same age may rarely use the internet, struggle with passwords or prefer a telephone conversation. The same variation exists at 30.

Digital capability varies between people, not simply between age groups. Age UK has highlighted the importance of listening to what older people want, keeping human alternatives available and helping people build confidence on their own terms.

That is a better starting point than “teaching seniors about technology”. The goal is helping people develop the digital skills they want, while preserving a genuine choice not to use a particular service.

AI could be extraordinarily useful

This is not an anti-AI article. Artificial intelligence could become one of the most useful technologies ever developed for healthy ageing if it is used carefully and designed around real needs.

Health trendsHelp make sense of activity, sleep, weight, blood pressure or heart-rhythm information where a validated system and appropriate clinical context exist.
Mobility and fallsIdentify a possible fall or a meaningful change in mobility, without treating an alert as a diagnosis or certainty.
Medication organisationOffer reminders and practical help with schedules, while leaving clinical decisions to the appropriate professional.
AccessibilityUse speech recognition, larger text, summaries and voice interfaces to make services easier to navigate.
Health informationExplain terminology and help somebody prepare useful questions for a healthcare professional.
Connection and cognitionSupport calendars, navigation, communication and everyday organisation for people who find them useful.

These are possible uses, not promises about what every AI system can do. A general-purpose chatbot is not automatically a medical device, and a confident answer is not the same as a reliable diagnosis.

Useful can become intrusive very quickly

Imagine a sensor notices that an 82-year-old has not left home today.

One interpretation is: “They may have fallen.” Another is: “They are watching the cricket.”

The technology sees data. It does not automatically understand a life.

Should it send a message? Contact a child? Contact a carer? Call emergency services? Wait? Do nothing?

Who decided: the software company, the family, the NHS, or the 82-year-old whose life is being monitored?

Technology intended to support independence might monitor movement, location, sleep, medication, heart rate, spending, driving, communication and activity in the home. That can be helpful. It can also reduce privacy and make somebody feel watched in their own life.

Support should not automatically mean surveillance.

Where a person can make the decision, they should be able to understand what is collected, why it is collected, who can see it, what the AI does with it, what happens when something unusual is detected and how to switch it off where appropriate. These are practical questions about control, not an attempt to provide legal advice.

“For your own good” is a dangerous phrase

Ageism can appear in technology as well as conversation.

Imagine a family installing monitoring throughout an older parent's house because “Mum's 80 now”. But Mum lives independently, understands the risks and does not want cameras or sensors.

Ageing does not automatically remove somebody's autonomy, privacy, preferences or right to take reasonable risks. Safeguarding concerns and questions about a person's ability to make a particular decision can make real situations more complicated. They still do not justify quietly treating every older person as incapable.

Technology should support independence where possible — not quietly remove it.

This is closely related to the concern behind elderspeak: adapting communication or support to an individual's needs is helpful; assuming reduced capability because of age is not.

AI can learn ageism too

AI systems learn patterns from data and from the choices people make when they design, train and deploy them. If the data under-represents older adults, contains stereotypes or reflects historic discrimination, an AI system can reproduce those problems.

An AI system does not have to “dislike old people” to produce age-biased results. Bias can emerge from:

DATA
Who was included, measured or left out?

DESIGN
Who tested the interface and whose needs were prioritised?

ASSUMPTIONS
What did the people building the system assume about age, ability or independence?

Researchers have described this as digital ageism. Possible examples include health models that perform differently across age groups, interfaces designed without older users, image generators that portray ageing through narrow stereotypes and recruitment systems that reflect existing age bias.

That does not mean every AI system is unsafe or unfair. It means testing across age groups, listening to affected people and checking outcomes are essential rather than optional extras.

The scam problem is becoming more convincing

AI can make fraud more sophisticated: cloned voices, deepfake video, convincing phishing, fake customer-service conversations, romance scams and investment adverts that appear to feature a trusted person.

This is not because older people are uniquely gullible. Scammers target people of any age because there is money or opportunity to steal. Older adults may sometimes face particular risks because they have accumulated savings or pensions, or because criminals deliberately target them.

The practical response is not suspicion of every new technology. It is a pause before acting on an unexpected request.

Stop. Check. Call.

If an unexpected investment opportunity, family emergency message, bank request, password request or payment instruction creates urgency:

STOP.
Do not act simply because the message tells you to hurry.

CHECK.
Look for another way to verify what you have received.

CALL.
Contact the person or organisation using a number you already know or have independently verified. Do not use contact details supplied in the suspicious message.

Use current NCSC guidance, report fraud through Action Fraud where appropriate, and check an investment firm with the FCA Firm Checker.

AI should identify itself

If somebody is communicating with a bank chatbot, healthcare chatbot, customer-service AI or the Age Watchers AI assistant, they should know that it is AI.

Disclosure is not a small design detail. It helps people judge how much confidence to place in a response and whether they want to continue. AI should not be designed to impersonate a healthcare professional or human coach.

The Age Watchers assistant is presented as an AI assistant and its interface explains that it provides general healthspan information only. It cannot diagnose or provide personalised medical advice. That boundary needs to remain clear as AI becomes more capable.

AI should not become your doctor

AI can help you ask better questions.

It can explain terminology, summarise information, help prepare questions, identify patterns for discussion and support habit formation. It should not pretend to be the person qualified to answer every question.

When something is medically concerning, ambiguous, high-risk or outside the system's competence, the useful response may be: speak to a human.

  • Possible heart-attack symptoms need emergency medical care.
  • Unexpected health deterioration needs an appropriate healthcare professional.
  • Complex medication questions belong with a GP, pharmacist or prescriber.
  • A possible scam belongs with the bank and the appropriate fraud-reporting route.

AI should know when not to keep talking. This is an editorial and design principle, not a claim that Age Watchers provides clinical triage.

Build with, not for

The most important principle is participation.

We should not sit in a room and decide, “Here is what older people need,” then build it without them. The people who use a service should help shape what gets measured, which reminders are useful, what becomes annoying, what AI should explain, what it should not touch, what can be shared and which privacy controls are understandable.

Age Watchers should follow the same principle. This is a product philosophy and an invitation to listen, not a claim that a formal user panel already exists.

OUR APPROACH TO AI

AI SHOULD IDENTIFY ITSELF.
You should know when you are interacting with AI.

YOUR DATA IS YOUR DATA.
We should be clear about what is collected, why and how it is used. The current privacy policy remains the source of truth.

AI SHOULD SUPPORT DECISIONS — NOT SECRETLY MAKE THEM.
Important health decisions require appropriate human involvement.

PEOPLE COME BEFORE AGE STEREOTYPES.
Do not assume capability based simply on someone's birthday.

USEFUL BEATS CLEVER.
Features should solve real problems rather than exist because they are fashionable.

HUMAN HELP STILL MATTERS.
Technology should know its limits.

BUILD WITH MEMBERS.
Member feedback should influence how Age Watchers develops.

What would you actually let AI do?

There is no single correct set of answers. That is the point.

  • Analyse your step trends?
  • Notice that your blood pressure has been rising?
  • Remind you to take medication?
  • Tell your family if you have not moved for several hours?
  • Listen for signs of cognitive decline in your speech?
  • Analyse your driving?
  • Monitor your spending for scams?
  • Book a GP appointment?
  • Share information with a doctor?

Different people have different boundaries. A person may welcome help with calendars and reject location monitoring. Somebody else may choose to share activity data with a family member but not a company. Healthy ageing technology should respect those differences.

Apply the Age Watchers framework

MEASURE.
AI can potentially help make sense of the measurements we choose to collect. But just because something can be measured does not mean it should be.

UNDERSTAND.
Know where information came from, what the system is doing and what its limitations are.

IMPROVE.
Technology should turn useful information into realistic actions, not simply generate more dashboards.

MAINTAIN.
Ask whether it helps maintain health, capability, independence, connection and choice — or whether managing the technology has become another burden.

Set your AI boundaries

Set your AI boundaries

Ask yourself:

  • What would I happily let AI help me with?
  • What information would I share?
  • What would I not share?
  • Who could see my health information?
  • When would I want a human instead?
  • Who should be contacted if technology detects something concerning?

Good technology should not simply know more about you. It should give you more control over your life.

Review your measurements

The real opportunity

AI could help make healthcare easier to understand, spot meaningful changes earlier, reduce administrative burden, support carers, make technology more accessible through natural conversation, connect communities and personalise healthy habits.

That is exciting. The objective is not to keep AI away from older people. It is to make sure older people are part of building what comes next.

The AI question is not: “Will older people be able to keep up?” Millions already are.

The more interesting question is:

Will the people building AI keep up with what older people actually want?

Healthy ageing should not be something technology does to us. It should remain something we participate in. If AI is going to become part of that journey, we should have a say in how it works.

MEASURE.
UNDERSTAND.
IMPROVE.
MAINTAIN.

One useful healthspan insight each day.

Follow Age Watchers for practical, evidence-based ideas built around: Measure. Understand. Improve. Maintain.

Follow Age Watchers on WhatsApp

Related reading

Sources / Further reading

AI, medical and privacy disclaimer

This article provides general educational information and is not individual medical, legal, financial or data-protection advice. Age Watchers AI features provide general healthspan information only; they do not diagnose, replace a healthcare professional, guarantee safety or detect fraud infallibly. Do not share sensitive personal or health information with an AI system unless you understand why it is requested, how it will be used and who may access it. Speak to an appropriate professional when a question is medically concerning, high-risk, ambiguous or outside the system's competence. For emergencies, use the appropriate emergency service.