AI tools and mental health support
If you visit FT online and sign up yourself and your students for free access to FT articles, you will find plenty of discussions on health and well-being, human relationships, and other topics relevant to your psychology course.
This month's FT online original article Use of artificial intelligence for mental health splits opinion, features Lucy Warwick-Ching, who discusses the use of AI tools for mental health support. These articles, along with the questions I provide, are an excellent way to give students relevant real-world examples that do not rely on research studies.
The use of technology in mental health
My article for a school DP psychology class relates this discussion to the use of technology in health and wellbeing. The use of chatbots to aid reflection through a form of “talking therapy” is becoming increasingly popular. Although the article focuses on 25–34-year-olds, the discussion is relevant to people of all ages. Students should click on the link to read the original FT article and then answer the questions posed in my schools psychology article. Suggested answers are in this blog.
Questions
- The author cites a 2025 report by Mental Health UK and Censuswide that found that 64 percent of 25–34-year-olds have used a chatbot for conversations about mental health or well-being. This is considerably higher than for any other age group. Why do you think this might be?
This age group is often still single and childless, so these young adults may be negotiating relationships, establishing or developing careers, and have more time to think about themselves and their own psychological health. The next age group is more likely to have family and childcare commitments and, therefore, less free time to focus on their own personal and health concerns.
Older people, particularly those over 65, may be less familiar or comfortable with AI systems and therefore less likely to use them for mental health discussions. They are also from a generation that may be more likely to trust a mental health professional than a chatbot. Adolescents and younger adults may have been omitted from the survey or may be less willing to admit to experiencing mental health or personal difficulties, or to using AI to discuss them, because of stigma.
You may want to draw students’ attention to how easily we can fall into stereotyping people of different ages when trying to explain patterns of behavior.
2. List at least three reasons given in the article for using AI to discuss stress or other mental health concerns.
- AI is seen as cheap, more confidential and less judgemental, instantly available and can provide expert advice. (According to Anna Maratos, head of psychotherapy for a London trust).
- Mental health systems have very long waiting lists. This takes strain off the services.
- Private mental health therapy is very expensive; this is free.
- AI doesn’t have to completely substitute for human care; it can be a first step for follow-up with a human clinician. It can also be used in combination with and alongside more traditional mental health care.
- It can be used between sessions with a human therapist.
- It may be used out of office hours.
3. What are the potential risks of using AI to help combat loneliness?
4. Research from Stanford University is cited, arguing that replacing human therapists with AI chatbots could be dangerous. Follow one of the two links provided and identify two potential dangers of using AI for mental health support.
- Even when a user’s thinking was the result of delusions, the AI system frequently validated these delusional beliefs, supporting their importance. This reinforces their psychological vulnerability and has even been linked to suicide risk.
- In a small number of cases, when users expressed violent thoughts, the AI chatbot encouraged self-harm, and only referred users to outside support half of the time.
- Some AI systems express stigma towards those with mental health conditions.
5. AI chatbots are criticized later in the article for showing a positive bias - being too willing to validate users’ beliefs rather than challenge them. Why might this be a problem, even when those beliefs are not delusional?
It can increase your own biases if you are told they are correct, whereas a human therapist can show you what to do about these biases.
People often go into therapy with a need to change their own attitudes and beliefs, but using an AI chatbot is more likely to reinforce these.
6. Do you think there is a future for using AI to discuss mental health concerns or other personal issues we are struggling with? If so, how might AI be used effectively and responsibly?
There are several options: completely replacing human therapists with AI support, except perhaps for those with the most serious mental health problems; using AI to point us towards the type of therapy and therapist we might benefit from; using AI in addition to sessions with a human therapist, to fill in the gaps in between therapy or to help us identify behaviour patterns and issues that we can then discuss with a professional therapist.
Student answers will vary. Most will probably see some sort of future for at least partial use of AI.
References
Criddle, C. (March 18 2026). AI chatbots often validate delusions and suicidal thoughts, study finds. FT online, https://www.ft.com/content/7f635a68-3b2a-4e4f-ae3d-926ff06ff068?syn-25a6b1a6=1
Moore, J., Grabb, D., Agnew, W., Klyman, K., Chancellor, S., Ong, D.C., & Haber, N. (2025). Expressing stigma and inappropriate responses prevents LLMs from safely replacing mental health providers. Proceedings of the 2025 ACM Conference on Fairness, Accountability, and Transparency (FAccT '25). Association for Computing Machinery, New York, NY, USA, 599–627. https://doi.org/10.1145/3715275.3732039
Warwick-Ching, L. (30 June 2026). Use of artificial intelligence for mental health splits opinion. FT online https://www.ft.com/content/a4ee9f36-74a5-4eac-8292-be2da893ae48?syn-25a6b1a6=1