AI mental health apps and chatbots: what they can and can't do

In short: AI chatbots and mental health apps can offer general information and self-help exercises, but they aren’t a replacement for a registered practitioner and shouldn’t be relied on in a crisis. Before you use one, check whether it’s regulated in the UK and how it handles your data.

More people are trying AI chatbots and mental health apps. This guide explains what they can help with, where the risks are, how to check whether an app is regulated in the UK, and why a registered human practitioner matters.

Uses your browser’s own voice. Nothing is sent to us.

Please read. This is general information from us, not medical, legal or professional advice. We’re a directory of registered practitioners: we don’t make, sell or review apps, and we don’t recommend any app or practitioner. We don’t offer an AI therapist or chatbot counselling. Links were checked on the official sites on 21 October 2026.

If you need help now

Call 999 or go to A&E if life is at risk. Ring Samaritans free on 116 123 (24 hours). Call NHS 111 and choose option 2 for urgent mental health help (England and Wales; in Scotland call NHS 24 on 111). Text SHOUT to 85258. Under 18? Call Childline free on 0800 1111.

What AI apps and chatbots can do

Used carefully, some digital tools can be a useful extra. Depending on the app, they may:

  • give general information about mental health and where to get help
  • offer structured exercises, diaries, reminders or self-help courses you can use at any time
  • help you put your thoughts into words before you speak to a GP or a practitioner

Some digital therapies are used in the NHS with support from a trained practitioner. The National Institute for Health and Care Excellence (NICE) has published early value assessment guidance on digitally enabled therapies for adults with depression and digitally enabled therapies for adults with anxiety disorders. It says some named programmes can be used in NHS Talking Therapies services while more evidence is gathered, once they have the right approvals, and that others should only be used in research. That is guided, assessed use inside a service, not the same as a general chatbot.

What they can't do

  • They aren't a person. A chatbot generates likely-sounding text. It can sound caring and confident and still be wrong.
  • They can make things up. AI tools can invent facts, sources or advice. Don’t rely on one for a diagnosis or for decisions about medication.
  • They may not handle a crisis safely. An app may miss that someone is at risk, or respond in an unhelpful way.
  • They can’t take professional responsibility. There is no register entry, no professional code of conduct and no fitness-to-practise process for a chatbot.

The main risks

Privacy

Information about your mental and physical health is special category data under UK data protection law, which means it needs extra protection. Some apps store what you type, share it with other companies or use it to train AI. Read the privacy policy before you type anything personal, and avoid entering names or details that could identify you or other people.

Crisis handling

If you are thinking about ending your life, harming yourself or someone else, or you are not safe, please contact a person straight away using the numbers in the box above, rather than an app.

Regulation

The Medicines and Healthcare products Regulatory Agency (MHRA) has published guidance on digital mental health technology. It explains that a product intended for a medical purpose, such as diagnosing or treating a mental health condition, with sufficiently complex functionality, must meet medical device rules. Many apps describe themselves as “wellbeing” tools, so they are not regulated as medical devices and haven’t had to show they are safe and effective for treatment.

No professional accountability

If you see a registered practitioner, you can check their register entry and raise a concern with their regulator or register if something goes wrong. That route doesn’t exist for a chatbot.

How to check if an app is regulated

  1. Read what it says it is for. The app store page or the app’s own information page should explain its intended purpose.
  2. Look for a UKCA or CE mark. The MHRA says a medical device app should show its UKCA or CE marking on the app store or its information page. See the MHRA’s advice on software, apps and AI.
  3. Search the MHRA register. The Public Access Registration Database (PARD) lets you search registered medical devices and manufacturers. The MHRA says registration is not an approval, certification or endorsement.
  4. Check NICE guidance. Search NICE for the app’s name to see whether it has been assessed, and what NICE said about how it should be used.
  5. Report problems. If a medical device app gives wrong results or you have a safety concern, you can report it to the MHRA through the Yellow Card scheme.

Questions to ask before you use one

  • Who makes it, and where are they based?
  • What does it say it is for, and does it show a UKCA or CE mark if it claims to treat a condition?
  • Where does what I type go, how long is it kept, and is it used to train AI?
  • What does it do if someone says they are at risk?
  • Is there an age limit, and is it suitable for me?
  • Is there a real person I can contact, and how do I complain?

Why a registered human matters

A registered practitioner has met their register’s standards for training and conduct, can respond to you as a whole person, can notice and respond to risk, and is answerable to their register or regulator. You can check that for yourself before you book. Our guides explain how to check a register and how to choose a therapist, and you can search registered practitioners on Find. A listing isn’t a recommendation, so please check the register yourself.

If cost is a barrier, see free and low-cost therapy and finding NHS support. Your GP can also help you work out what might suit you.

Common questions

Is an AI chatbot the same as therapy?

No. A chatbot is software that predicts text. It is not a person, it is not on a professional register, and it cannot take professional responsibility for your care. Some apps use therapy-style exercises, but that is different from working with a registered practitioner.

Are mental health apps regulated in the UK?

Some are. The Medicines and Healthcare products Regulatory Agency (MHRA) says digital mental health technology that is intended to diagnose, prevent or treat a condition, and has enough complexity, must meet medical device rules. Many wellbeing apps are not intended for a medical purpose, so they are not regulated as medical devices.

How can I tell if a mental health app is a regulated medical device?

Look on the app store page or the app's own information page for a UKCA or CE mark and what the app is intended for. You can also search the MHRA's Public Access Registration Database. Registration there is not an approval or endorsement by the MHRA.

Is it safe to tell an AI app about my mental health?

Information about your mental health is special category data under UK data protection law, which needs extra protection. Before you type anything personal, read the privacy policy: who runs the app, where your words go, how long they are kept, and whether they are used to train AI.

What should I do if I am in crisis?

Please contact a person, not an app. Call 999 if life is at risk, ring Samaritans free on 116 123 at any time, call NHS 111, or text SHOUT to 85258. If you are under 18, you can call Childline on 0800 1111.

Sources, checked 21 October 2026: MHRA: digital mental health technology · MHRA: software and AI as a medical device · MHRA PARD · MHRA Yellow Card: software, apps and AI · NICE HTG675 · NICE HTG676 · ICO: special category data. Therapist Circle is independent and not affiliated with the NHS, the MHRA, NICE or any professional or regulatory body.

Are you a registered practitioner? List your practice free in our free UK therapist directory.

Find a therapist · All guides · Useful links