This post reflects on reporting by The Guardian (December 2025) examining how some teenagers are using AI chatbots as a form of emotional support.
Recent reporting by The Guardian has drawn attention to a growing number of teenagers turning to AI chatbots when they feel distressed, lonely, or overwhelmed. The coverage has raised understandable questions about safety, effectiveness, and what this emerging pattern might mean for mental health services working with children and young people.
This is not a therapy study, and it does not offer clinical outcomes or treatment guidance. However, it does highlight something important for practice: some young people are already seeking emotional support in digital spaces outside traditional services.
This post does not take a position on whether AI should or should not be used for mental health support. Instead, it explores what this behaviour might tell us about young people’s needs — and what it may invite therapists to stay curious about in everyday clinical work.
What the reporting highlights
Recent coverage describes some teenagers using AI chatbots when they feel distressed, lonely, or overwhelmed. Young people describe these tools as:
- Always available
- Non-judgemental
- Easy to talk to
- Free from perceived pressure or scrutiny
Alongside this, clinicians and researchers have voiced concerns — particularly around safety, boundaries, accuracy of responses, and the absence of human understanding when distress or risk is high. Importantly, these reports are descriptive, not evaluative. They don’t tell us whether chatbot use improves mental health outcomes, who it may help, or who it may place at greater risk. What they do show is that some young people are already using these tools — quietly, independently, and often outside of professional awareness.
What this doesn’t tell us
It’s worth being clear about the limits here.
These reports don’t tell us:
- Whether AI chatbots are effective mental health interventions
- How young people interpret or internalise chatbot responses
- How use differs by age, vulnerability, or level of risk
- Whether chatbot use complements, replaces, or delays help-seeking
Without this information, it would be premature — and unhelpful — to draw firm conclusions. But uncertainty doesn’t mean irrelevance. It means we need to think carefully about how this shows up in the wider context of young people’s lives.
A clinical lens: what might this be meeting for young people?
Rather than focusing on the technology itself, it may be more helpful to ask: what needs are being met here?
Many of the reasons young people give for using chatbots echo themes we already recognise in therapy:
- Wanting to feel heard without fear of judgement
- Needing immediate support outside service hours
- Struggling to put feelings into words with another person
- Feeling unsure whether their distress is “enough” to bring to an adult
Seen this way, AI use may not reflect a rejection of therapy — but a response to gaps in availability, accessibility, or perceived safety. For some young people, it may feel easier to start talking to something that doesn’t look back, interrupt, or ask follow-up questions they aren’t ready for.
That doesn’t make it a substitute for therapy. But it does tell us something about how young people seek support when they feel unsure, exposed, or overwhelmed.
Bringing this into therapeutic thinking
In practice, this isn’t about screening for AI use or treating it as a risk marker. It’s about curiosity and formulation.
If a young person mentions using an AI chatbot, it may be helpful to explore:
- What they find helpful about it
- What they turn to it for — and when
- How it makes them feel before, during, and after use
- What they don’t feel able to bring into other relationships yet
These conversations don’t need to be alarmist or directive. They can sit alongside broader exploration of coping, help-seeking, and trust — particularly for young people who feel unsure about being understood.
From a CBT perspective, this also raises questions about:
- How young people make sense of advice or reassurance
- How they test or apply what they’re told
- Whether responses reduce distress in the moment or shape longer-term beliefs
Again, we don’t yet have answers — but noticing these processes matters.
Holding boundaries without shutting down curiosity
Concerns about safety, particularly around risk and the absence of human judgement, are valid and important. At the same time, responding too quickly with prohibition or alarm risks closing down conversations that may already feel difficult for young people to have.
A reflective stance allows therapists to:
- Acknowledge uncertainty honestly
- Hold boundaries clearly and proportionately
- Stay curious about meaning rather than focusing solely on behaviour
This is often where therapeutic work is most helpful — not in offering certainty, but in supporting understanding.
A final reflection
Rather than drawing conclusions, this reporting leaves us with important questions for practice:
- What does young people’s use of AI tools tell us about when support feels accessible or safe?
- How might anonymity and immediacy shape help-seeking behaviour?
- What needs are being met — and which are not — when support comes from a non-human source?
- How can therapists remain open to these conversations without overreacting or minimising risk?
- What would it look like to hold technology use within formulation, rather than treating it as a problem to solve?
These are questions the field does not yet have clear answers to — but they are worth holding as this area continues to develop.