AI and Mental Health

AI is already a part of our everyday lives, it’s changed and is changing the way we study, plan our lives, keep motivated and encouraged and access emotional support. 

Using AI to support mental health is less clear, in fact it’s a complicated picture; this article will explore some of the important questions about AI and mental health including - dependency, safety and some of the issues of interacting with something that seems both professional and human.  

What do we mean by AI and avatars in mental health?

AI chatbots are online ‘tools’ that respond to questions posed, which can be typed or spoken; they are included on many websites, apps and can be great for asking questions like ‘What can the Hive help me with?’ or ‘Does my policy cover for windscreen damage?’ when it comes to mental health AI chatbots can be a little tricker.

Chatbots often use avatars as visual or voice-based characters that may look, sound or behave like a supportive person and some are specifically designed for wellbeing; they can be emotionally supportive, reassuring and comforting. 

It’s very important to remember that AI chatbots are not a mental health service, they are designed to give positive, supportive and affirming responses which do not acknowledge the clinical aspects of supporting mental health appropriately. 

But I want to use AI for emotional support

AI is convenient, it’s quick, available 24/7 and non-judgemental; therefore, it makes sense you’d be drawn to using it and let’s be honest, talking to a person aloud can feel intimidating at times. Plus, AI helps you in an instant, no referral, no waiting lists – support right away. 

If you are looking for some quick tips on how to deal with a panic attack in the short term, AI can be an answer; although AI should not replace professional human support or be a longer-term mental health strategy.

What are the potential benefits?

There are benefits to using AI for wellbeing support. For example, AI can help put thoughts into words, give journal prompts, help prepare for a difficult conversation with a tutor. AI provides some really helpful psychoeducation. It can clearly explain concepts like anxiety, grounding techniques and procrastination. 

AI also helps reduce barriers to seeking help; you might know you need support but may feel hesitant about contacting the Centre for Wellbeing. AI can be the bridge to help you build your confidence prior to meeting a mental health professional; however, chatbots or AI tools cannot replace professional support. 

Are there any drawbacks or risks?

AI can be really useful, but it is not always right. Sometimes it gives answers that sound very confident, even when the information is wrong, unhelpful or even unsafe.

It can feel like AI understands what you are going through, especially when it responds in a kind or empathic way. But AI does not really understand feelings, relationships or risk in the way a professional does. So, AI can be a helpful starting point for information or ideas, but it should not replace talking to a qualified professional, especially if you are worried about your wellbeing, safety or mental health.

It may also encourage dependency. If you start relying on an AI companion or avatar, there is a risk of disengagement with the wider world and isolation. 

AI may miss risk. A chatbot may not reliably recognise the nuanced nature of risk or even subtle risk indicators in the same way a mental health professional would. Under these circumstances, using AI could be more unhelpful than you might expect. 

Can AI replace counselling or therapy?

The answer should be clear: no. AI can support the space around therapy, but it should not be mistaken for therapy itself.

For example, it may help you prepare for counselling, reflect after a session, or write down what you want to discuss. But it should not be used as the only source of support especially if you are at risk, deteriorating or struggling to function.

What about privacy and confidentiality?

When you see a counsellor, there are strict confidentiality protocols and standards in place, for AI tools this is not the same and may not necessarily offer the same level of confidentiality. Therefore, you should avoid giving personal information, especially highly sensitive information unless you fully understand how the AI platform manages your data. 

Before using AI for something personal, ask yourself: Would I be comfortable if this information was leaked, or connected back to me? If your answer is no, then it’s best not to use AI. 

Could AI make mental health worse?

Unfortunately, for some people, yes it can. AI may increase rumination or support avoidance of seeking professional help. AI can sometimes create false reassurance, increase distress if the response is confusing or unsafe, or make loneliness worse if it starts to replace real human connection.

When might AI be useful for me?

AI tools, chatbots or avatars definitely have their benefits when it comes to mental health. Safer uses of AI include asking for general information about stress, sleep, anxiety or wellbeing. Creating a study plan when feeling overwhelmed, practising how to start a conversation with a GP, or counselling service.

When should I not rely on AI?

AI should not be relied on alone; it is better to source a range of support options. AI should absolutely not be used when:

• You feel at risk of harming yourself or someone else.
• Feel unable to keep yourself safe.
• Are experiencing abuse, coercion or exploitation.
• Are hearing voices, feeling paranoid, or losing touch with reality.
• Are severely restricting food, purging, misusing substances or taking significant risks with your health.
• Feel you cannot function day to day.
• Need urgent medical or safeguarding advice.
• Are being encouraged by the AI to keep secrets, isolate, avoid real help or take risky action.

In these situations, you should contact campus safety, emergency services, NHS 111, your GP, the centre for wellbeing or trusted people around you. 

Is there a way to use AI safely for mental health support? 

A safer option is to use support that the Wisdom App, which is part of the wellbeing support the university provide. The Wisdom app gives free access to wellbeing resources, mood tracking, mini health checks, guided content and support around mental, physical, social, financial and legal wellbeing. The app also offers free live chat and video call options with counsellors 24/7.  

You might use the Wisdom app to check in with how you are feeling, look at wellbeing resources or reach out for support when you are finding things difficult. It can be particularly helpful outside normal university opening hours, when you may want some immediate support or guidance.

What is the final advice? 

A sensible approach is to treat AI like a tool in your wellbeing toolkit, not the whole toolkit. It can be useful, but your mental health deserves more than an automated, machine-learnt response.

The key is to use AI with awareness. Notice how it affects you. Does it help you feel clearer, calmer and more able to act? If yes, then continue, if you it keep you stuck, isolated, overthinking or avoiding real support? Then contact the Centre for Wellbeing.