People Are Already Using AI for Mental Health. This Changes the Debate.

6 days ago
3 min read

A recent survey from Mental Health UK found that more than one in three adults (37%) have used an AI chatbot to support their mental health or wellbeing.

We are no longer debating whether AI will play a role in mental health support; it already does.

The real question now is: what kind of AI are people talking to?

The reality: people are turning to AI first

There are many reasons why people turn to AI tools when they are struggling.

· AI is available instantly

· It doesn’t judge.

· It’s private.

· And for many people, it feels easier to start a conversation with a machine than another person.

At a time when mental health services are stretched and waiting lists remain long, it is not surprising that people are exploring alternatives. Research and professional bodies have warned that overstretched services are one factor pushing people toward AI chatbots for support.

In many cases this initial interaction is simply a starting point. People are looking for reassurance, reflection or somewhere to organise their thoughts before seeking further help.

That potential is why AI in mental health is attracting so much attention. But it is also why it demands careful design.

Not all AI is built for mental health

Recent investigations have highlighted what can go wrong when general-purpose AI tools are used in sensitive areas like mental wellbeing.

A Guardian investigation prompted the UK mental health charity Mind to launch a year-long inquiry into the impact of AI on mental health, after examples emerged of misleading and potentially harmful advice generated by AI systems.

The concern is not that AI should never be used. The concern is that general-purpose AI systems are not designed as mental health tools. They are trained to generate plausible responses to almost any question, not to provide structured support or recognise vulnerability.

In mental health, that difference matters.

Someone seeking support should not have to question whether the advice they receive is safe or appropriate. Yet experts warn that publicly available AI tools do not always have the safeguards required for people in distress. If people are going to use AI for mental health, then the AI they encounter needs to be designed specifically for that context.

The question is no longer whether AI should be involved in mental health support. The focus should be on how that support is delivered responsibly.

That means systems which are:

grounded in psychological research
designed with clinical oversight
built with strong safeguards and escalation pathways
transparent about what they can and cannot do

AI should never replace clinicians. But it can act as a front door to support: a place where someone can begin to talk, reflect, and be guided toward the right next step.

When designed properly, AI can help remove some of the barriers that prevent people from seeking help in the first place.

Why the design of these systems matters

This is one of the key principles behind Aria.

Aria was built specifically to support everyday mental wellbeing conversations. It draws on established psychological approaches and is designed to work alongside existing healthcare services rather than replace them.

The aim is not to provide diagnosis or therapy. It is to offer a safe, structured and confidential space where people can begin to talk about how they are feeling and be guided toward appropriate support.

For some people, talking to an AI may be the first step they are comfortable taking. If that interaction is well designed, it can help them reflect, regulate and, where appropriate, seek further help.

But if that first interaction is poorly designed, it risks doing the opposite.