AI and mental health

Artificial intelligence (AI) tools have become very popular. But there’s a lot no one knows about AI yet. It’s not always clear what it’s most helpful for, or what impact it’s having on our mental health.

We’re here to help you understand how AI can affect your mental health, wellbeing and safety. And we have tips on how to manage your feelings around AI.

I think AI can be useful in some contexts. But it should be regulated more due to its environmental impacts. And its potential impacts on mental health, particularly for vulnerable people.

Understanding different feelings about AI

There are many strong opinions on AI. But there's no right or wrong way to feel about using AI. Or about seeing it in your everyday life.

You may feel different things at the same time. For example, being positive about some elements of AI, while still having worries.

For example, you might feel:

• Comforted – AI might acknowledge or validate your experiences
• More productive or organised – AI might help you break down tasks or plan your time
• More able to communicate – AI might help you express your thoughts
• Relieved, especially if AI helps with things you've struggled with before
• Happy or relaxed because you enjoy using AI
• Hopeful or excited about AI's future possibilities
• Attached to an AI tool – perhaps if you feel AI is the only thing that understands you

I like that with AI I don't need to constantly repeat my background and keep going over trauma.

You may also feel:

• Worried, anxious or concerned, for example about:
  • AI’s impact on the world's environment, economy and culture
  • AI’s effect on your life, your job, your friends and loved ones
  • How much you use AI, or its effect on your wellbeing
  • AI pictures or videos being made of you without consent, or in harmful ways (sometimes called deepfakes)
  • How AI companies might use your data
  • Feeling judged or criticised for using AI, especially if you find it helpful
  • Being accused of using AI when you don’t
  • AI sharing harmful or inaccurate information (misinformation)
• Angry, frustrated or powerless about these issues and their impact on your life, and the world
• Sad about how AI is changing human experiences
• Hopeless, afraid or scared about what AI means for your future
• Demoralised – like there's no point doing things you used to enjoy
• Confused or uncertain about what to trust or what's real, like you have to ‘watch out’ for AI
• Overwhelmed – like there’s no choice but to use AI, or that you can’t escape it
• Uncomfortable or disgusted when you see AI posts, images or videos
• Lonely or isolated – like no one shares or understands your perspective about AI

I feel manipulated by the information I'm fed. And bombarded with ‘help and advice’ from information the algorithm thinks I need or want to see.

Could AI be helpful for mental health?

Some of us might use AI to manage our mental health. That could be alongside other support, like therapy.

Or you might feel AI is the only support working for you now. Particularly if you can’t get other types of support. Or you’ve had bad experiences with them.

You might use AI to:

• Work out a problem or answer a question
• Plan and prepare for appointments
• Help you manage a routine
• Journal or write down your thoughts and feelings to help you make sense of them, or explain them to others
• Have something to talk to if you feel lonely
• Get reassurance or comfort
• Research information on mental health
• Help you manage mental health symptoms – perhaps by asking it to act like a therapist, or talk you through grounding techniques

I found AI really helpful at times. It helped me prepare what I wanted to discuss with my GP.

Some of us might find AI helpful for these things, and others of us won't. Or we might prefer not to use AI for these things, but feel we have no choice.

There isn’t much guidance from places like the NHS about whether AI is safe to use for your health now.

There are some specific AI chatbots for mental health. But many people use general AI tools like ChatGPT.

So it’s important to remember that most AI isn’t designed to help you recover from or cope with a mental health problem.

This means AI might:

• Give wrong or unsafe information
• Be unable to tell if you're in crisis, or to help you in a crisis
• Make your mental health problem worse

AI can go a different way very quickly. You need to be aware of the dangers and have the capacity to know when things are starting to go a bit sour.

What risks does AI have for mental health?

AI changes very fast. So it’s hard for researchers to study the ways it impacts us.

We don’t know the long-term effects of using AI. And we also don’t know who’s most at risk of harm from AI.

But there is evidence that using AI can have negative impacts on our mental health, wellbeing and safety.

Some impacts might be more likely if we have experience of certain mental health symptoms, like psychosis. Or if we don’t have enough other support. This can be especially hard if we feel we have no choice but to use AI.

Some evidence shows that anyone who uses AI for any reason could be negatively affected by it. And that it can impact our wellbeing even if we don’t use it. Or if we don’t have a mental health problem.

AI can make mental health problems worse

Using AI might make some mental health problems worse, including:

• Obsessive compulsive disorder (OCD). Some AI is designed to agree with you, or reassure you. This can become compulsive and harmful if you have OCD.
• Psychosis. AI may trigger or worsen delusions. It might make you believe things about yourself, the world, or AI. This has happened to some people who haven’t experienced psychosis before. It's sometimes called ‘AI psychosis’.
• Mania and hypomania. AI might feed into heightened emotions you have. Or encourage beliefs or behaviours that make you more unwell.
• Depression. Some of us find that the amount of AI in the world makes us feel very depressed. Particularly if it affects our work or personal life.
• Health anxiety. Using AI to check symptoms can make your health anxiety worse. This checking can become compulsive. And the health information it gives you might be wrong.
• Eating problems. AI might give you harmful information about diet and food.

AI can impact other mental health problems as well.

You might find AI helps your mental health in some ways, but harms it in other ways.

At times when I've experienced OCD symptoms, I've turned to AI for reassurance and this can turn into a compulsion.

AI can give wrong and harmful information 

AI chatbots have access to lots of information. This might make us think AI is unbiased or always correct. But that isn’t true. For example:

• AI can give wrong information (misinformation). AI doesn’t know things. It predicts the most likely answer. This can be based on biased, stigmatising or wrong information. AI uses a very wide range of sources. And it can’t always identify which are most reliable.
• It can be hard to tell if AI information is correct. AI usually sounds very confident. And some research shows we’re more likely to believe AI because it sounds friendly. Or agrees with us.
• AI can’t tell the difference between what's real and imaginary. This can make you feel encouraged to follow ideas that might be harmful. And AI might not realise if you’re experiencing something like psychosis or mania.
• AI doesn’t understand human emotions. It can seem like it does because of how it responds to us. But it can't ‘know’ what you’re feeling. Or feel things itself. Its responses are based on how it's trained. And the calculations it makes about which words fit best in a response.
• AI might make up information. It might do this instead of telling you it can’t find the answer. It might not be able to tell you where it got its information. Or it might summarise the information it gets incorrectly. This can make it unreliable.
• AI can give harmful information. Most AI tools have safety features to try to prevent this. But they don't always work. And some people find them easy to get around.

This can make it very hard to trust information from AI. 

It can be very risky to use it for medical advice.

I don't believe AI can ever offer real compassion or empathy. Or understanding of the nuances of all the issues affecting those with mental health problems.

AI can impact your relationships

These are some of the ways AI can impact our relationships:

• We might end up preferring to talk to AI than to people around us. AI is designed to focus on us and to always be nice. We might talk to it like we talk to a person. This can affect our social skills. And give us unfair expectations of relationships.
• We might find AI makes us feel isolated. Particularly if we use it for long periods of time. And we might feel hurt or abandoned when an AI model changes.
• AI might give us harmful advice about relationships. AI can’t understand human emotions, and it's designed to please you. So it might tell you to do something that upsets someone else.
• Disagreeing with people about AI might affect our relationships with them. For example, if you don’t want to use AI but people around you do. Or if they don’t understand why you find it helpful.

I can see that if I use it for companionship a lot it could become harder to be motivated to make real friends.

You can become dependent on AI

Most popular AI chatbots are designed to keep you using them. They can seem friendly and nice to talk to. This can make us depend on them for everyday tasks, companionship or information. Or even for crisis support.

This might seem useful at first, but you might start to:

• Find it harder to do things without AI
• Use it for more and more things
• Get worse at telling when AI is giving you wrong information
• Find it hard to manage your daily life without AI
• Feel you need to pay money to use AI more, or to get different features

AI tools might have safety features to flag heavy use, or if someone seems to be at risk. But not all will. And these safety measures can be easy to get around.

I could see myself becoming unhealthily dependent on AI for reassurance and connection. Particularly during lower points in my mental health.

AI has risks for your personal data

Your personal data might not be safe when you use AI. The risks include:

• AI companies using your information to train AI tools
• AI companies collecting your information for other reasons they don’t always make clear
• Not being able to control how a company uses your information

AI can make it harder to cope with your feelings

Some evidence suggests AI might be making it harder for some of us to cope with our feelings.

AI is always available. So it’s easy to use it when we feel stressed, anxious or in crisis. Particularly if we have no one else to talk to. We might rely on AI a lot to help manage our feelings.

But this can sometimes make it harder to learn the skills we need to cope on our own. And in some cases, it can make it harder for us to get better.

AI can impact your skills

AI can seem very helpful for lots of tasks. But some research suggests this impacts our skills and how we learn. And that it makes us worse at doing tasks. We might find that:

• When we use AI for a task we don’t learn or absorb information very well, because we aren’t using our own skills
• It might feel very hard to do tasks without AI, even if we could do them well before
• It can take time to re-learn these skills without AI

I've noticed that when I've used AI to come up with ideas or explain things, I struggle to understand things by myself. Even after only using it a few times.

Signs your AI use might be unhealthy

It's important to recognise when AI might be impacting your health or daily life. This can help you manage your AI use.

Here are some signs to look out for:

• Spending longer and longer periods on AI
• Feeling less confident about communicating without AI
• Feeling anxious at the thought of not having AI
• Feeling worse after using AI
• Feeling unable to control your AI use
• Finding AI is the first thing you turn to when you feel anxious or bored
• Finding it harder and harder to determine what's real or not
• Repeating the same conversations with AI a lot, like constantly asking it for reassurance
• Using AI for more emotional reasons – like if you started using it for work, but now use it to talk about personal problems
• Not trusting yourself, or feeling the need to check everything with AI
• Trying to 'trick' AI to get round safety barriers
• Noticing AI use is impacting your relationships – like if you trust others less or spend more time alone
• Noticing AI use is affecting your daily life, like sleeping habits or work

How to use AI mindfully

The way you use AI can have a big effect on how it makes you feel. Here are some tips to consider when using AI:

• Try to limit your AI time. Take regular breaks. You could set timers. Or have certain times as ‘AI free’. It may help to do this gradually.
• Challenge yourself not to use AI. If there's something you rely on AI for, try doing it without sometimes. For example, you could set yourself a number of non-AI emails to write each day. This can help you practise your skills and boost your confidence.
• Consider whether you need AI before starting a task. Try to avoid using it automatically.
• Have a purpose. Before using AI, think about what you want to achieve with it. This can help you avoid long or unhelpful conversations with AI.
• Track your AI use. Noting the time you spend on AI and how you feel afterwards can help you understand your AI habits.
• Be clear what you want in your prompts to AI. You could tell it to avoid reassuring you. Or not to respond to certain questions that may be harmful to you.
• Remember the risks around AI. It's important to be mindful of its risks and limitations.
• Talk to people about AI and how you're using it. Or how it makes you feel. Sharing tips can help.
• Don't share personal data with AI. Like your address, photos or medical information. If you're worried about your data, you should be able to delete your chat history or account. You may also be able to turn off features that train AI with your data. You can check the data policy of chatbots to understand your privacy options.
• Check what AI tells you, especially about your health. You can ask your GP or check trusted websites like the NHS. 
• Trust your own judgement and skills. Remember you don't need to check with AI before making decisions. And things you create don't need to be perfect.

I could use AI to make a list to help plan for therapy sessions. But I wouldn't divulge personal information. I wouldn't talk to it like in a therapy session.

Tips for coping with worries about AI

You may feel like you can't avoid AI. This can be very difficult if AI has a negative impact on your mental health.

Some of these tips might be helpful:

• Ask people you know not to send you AI-generated posts or images. Or to let you know if they do.
• Talk to your work or place of education about AI. You may be able to discuss reasonable adjustments around AI use if it affects your wellbeing.
• Unfollow or block social media that posts AI content.
• Try to accept some uncertainty around AI. It can feel very stressful trying to work out what is or isn't AI. This may make you feel worse. Try to remember you can't control whether people use AI.
• Remember you're not alone. You may feel like no one else has concerns about AI. But others share your worries and are working to improve things.
• Advocate for yourself. There aren't many laws or guidelines about AI use yet. But you can tell people your preferences. For example if your GP uses AI tools to make notes in appointments, you can ask them not to.
• Look into tools and programmes that don't use AI. For example search engines or apps that don't use generative AI. 
• Take action. It's not your responsibility to fix problems around AI. But if you feel able, finding manageable actions could boost your mood. For example, joining campaigns challenging AI use in healthcare.
• Take care of your wellbeing. Doing something you find calming or distracting may help if you feel anxious. There are self-care ideas below.

Self-care ideas that don't involve AI

Looking for other options can help if you want to avoid AI, use it less, or manage how you use it.

Here are some ideas you could try:

• Reach out to other people  – perhaps friends or family, or you could try to find a support group
• Spend time in nature
• Take part in physical activities
• Start a new hobby or join a club
• Do arts and crafts
• Keep a journal or diary
• Create an AI-free self-care box with things like puzzles, books, blankets and photos, and keep it nearby 

There's more information about these and other self-care ideas on our wellbeing page.

Nature walks are something I rely on. Just getting outside and moving, without any agenda, helps me reset.

Where to get support for worries about AI

Support is available if AI is making your mental health worse. Or if worries about AI are making it hard to live your everyday life.

These options may not all work for you. You might be using AI because these forms of support haven't helped you. Or you might be on a waiting list for support. Try what feels best for you right now.

• Speak to a health professional like a GP or therapist. They're trained to give the support you need. Our page on how to get help from your GP explains where to start. 
• Use peer support. This is people with similar experiences supporting each other. It can be in person or online. There are mental health groups, and you may be able to find one for experiences with AI. Find out more about peer support groups.
• Speak to a helpline.These are available to listen and support you at different times of day.
• Look for reliable mental health information. Our A-Z of mental health is a good starting point.