Mental health charities support people from all walks of life, everyday, on their journey to better health and wellbeing. Our workforces are in person, over the phone, online, supporting people when they need it most. So where does AI come in?  

In the first blog in our series on AI in mental health charities, Lee Hall from Everyturn Mental Health told us about the importance of governance and maintaining boundaries around how AI is used – not compromising an organisation’s values or compassion. 

For the second blog in our series on AI in mental health charities, we spoke to our member Anxiety UK, to find out what they’ve learnt so far about AI usage within their charity.  

Anxiety UK is a national charity run by and for people with anxiety, and supports people through a 24/7 information and self-care line, talking therapies services, support groups and more. 

How are you using AI at your mental health charity and what have you learnt? 

At Anxiety UK, we’re using AI in clearly defined, “behind the scenes” ways to support our staff and volunteers rather than replace them. We use it to help create and refine content for our website, promotional materials and social media, and to update existing documents so that they read with a tone that resonates with the specific audience we’re trying to reach. This has saved us huge amounts of time, allowing staff to focus on other tasks. The AI package we use is now seen as an integral part of our toolkit: we use it predominantly both to form the basis of documents that we need to produce and to finesse and polish them.  

We also use AI extensively to acquire knowledge – for example, to scan and summarise research and undertake literature searches so we can better understand the evidence base and what the literature says about topics relevant to anxiety. In fact, we used our AI tool to assist with the creation of this very blog, which for us illustrates how AI can be a powerful timesaving tool for communication and research. 

How have you made sure that your use of AI tools is safe and ethical? 

We’ve taken a cautious, values-led approach. We do not use AI to replace therapeutic judgement, make clinical decisions, or provide direct one-to-one support to people in distress, and we never allow AI to operate without human oversight. We assess tools for data protection and security, avoid inputting identifiable personal information, and provide clear guidance and training for staff on what AI can and cannot be used for. We are also guided by our ethos as a user-led charity run by and for people with lived experience of anxiety, so we actively listen to our community’s views and concerns and factor these into our decisions.  

To get started, we identified experts in the field of AI who very kindly delivered a number of pro bono general AI training sessions for our small Anxiety UK team. We then drew on their expertise to identify a suitable AI package, asked our DPO to assess it in terms of data security and governance, and only then embedded it within the organisation’s IT infrastructure and trained staff in how to use it. At the same time, we updated our IT and Data Protection & Confidentiality policy to set out clearly how we use AI, the safeguards we have in place, and the limits we place on its use. 

What do you find promising about the future of AI in your work as a mental health charity? 

For us, one of the most promising aspects of AI is its ability to save time by helping us quickly create and refine high quality content and resources for people with anxiety. It can also rapidly analyse data and support informal literature searches, giving us a clearer understanding of particular areas of anxiety and helping us stay close to the evolving evidence base.  

Looking ahead, we’re excited by the idea of an AI “dashboard” or suite of tools that would allow us to generate images, videos and other materials from within the scientifically robust, academically focused AI environment we already use, rather than relying on multiple separate tools. Above all, what matters to us is fast access to accurate information drawn from credible, trustworthy sources so that anything we produce using AI remains reliable, evidence-informed and genuinely useful for the people we support. 

What are your biggest concerns going forward around the use of AI within mental health charity service provision? 

We have a number of concerns, particularly around how people may come to over-rely on technology. For example, for those living with social anxiety, heavy use of digital tools can contribute to avoidance of real-world contact, which risks exacerbating difficulties rather than helping them. Human, in-person interaction remains hugely important.  

We are also concerned that more people are using AI tools to “therapise themselves” without fully understanding the data privacy implications or the limits of AI when it comes to mental health support. Our sense is that this kind of AI help-seeking is likely to grow until the risks are better understood and/or meaningful safeguarding controls are in place for these interactions. In a somewhat ironic way, we suspect that this might ultimately prompt a renewed appreciation of something very traditional: the real, human therapist. 

Stay tuned for more blogs in this series, discussing the future of AI in mental health charities. 

To join the discussion, become a member of the Association of Mental Health Providers today.