A vast live experiment in young people's mental health is being conducted as they increasingly turn to AI companions for advice and assistance.
No one signed a waiver opting in to the trial, but its results are fast becoming a reality for our societies.
Perhaps it is unsurprising that “AI use is surging for mental health help among young people” , given the worldwide shortage of child mental health experts or tailored mental health programs for young people.
Whatever the reasons for the surge, what is certain is that millions are turning to AI for support, companionship and advice and the speed of uptake is far outpacing our ability to understand the effects and provide the required safeguards.
Safety, Trust and Human Connection
In a context like this, when technology moves faster than our understanding of its consequences, bringing the right people together becomes essential.
That's why the WHO Foundation is helping to bring young people, health experts, researchers, policymakers and technology leaders into the same conversation. At this year's World Health Summit and UN General Assembly, we explored both the extraordinary potential of AI to support young people's mental health and the urgent questions it raises about safety, trust and human connection.
These conversations are part of a wider effort to ensure that the future of AI and mental health is shaped not just by what's technologically possible, but by what's genuinely good for young people. Many contributors could see the potential for generative AI models to contribute to young people’s mental health. They are:
Easy to access - whereas finding a doctor can take time and several referrals
Promise anonymity - which appeals to users who feel hesitant about sharing information where they would be judged or stigmatized
Patient and accommodating - consultations with medical professionals tend to be brief - generative AI doesn’t look irritable or seem too rushed to listen
Friendly and approachable - often greeting users like an old friend, no jargon, no hierarchies
But there were also fears that we’re playing out this social experiment in real time, or as one contributor expressed it: “We are holding an open flame. Until the systems have more guidance, regulation and stewardship, [my feeling is] more trepidation than excitement.”
Questions include whether AI models were acceptable stand-ins for health professionals - and if, by asking a bot for advice on everyday problems - we risk young people finding it harder for them to develop resilience and the social skills they need to safeguard their future mental wellness.
Plus:
No responsibility or accountability - AI bots aren’t regulated by health authorities and their mental health advice is unmonitored, unchecked and unverified
Potential for increased isolation - in avoiding difficult conversations, young people are also missing out on human relationship-building
False friends - young people may not realize the friendly bot is not a human and it isn’t a real relationship. As one contributor put it: “You lose sight that you're not talking to a person.”
Economic model is built on keeping people online : AI bots are built to keep users engaged, not with the interests of the user at heart, but with the interests of the platform developers.
As the AI conversation progresses, the WHO Foundation will bring more news, insights and evidence about AI and the role of WHO and others in safeguarding our health on digital platforms.
The WHO Foundation co-hosted the event The Truth about Youth and Bots, A Health Approach to AI with the digital wellbeing program Ithra Sync on the sidelines of UNGA 81.
For more updates about AI and mental health follow the WHO Foundation on LinkedIn.
