Youth Adoption and Private Usage Patterns
Generative artificial intelligence has increasingly stepped into the personal lives of young people, serving as a informal mental health support system. According to survey data released by Canadian youth mental health network Jack.org, one in three young Canadians now relies on generative AI chatbots at least once a month for mental health support and emotional guidance.
The study highlights a significant degree of privacy surrounding these interactions. Nearly half of the respondents who utilize conversational agents for mental wellbeing reported that they have not disclosed their AI usage to family members, friends, or medical professionals. Among active users, 60 per cent utilize chatbots to navigate complex social situations, including strained family dynamics, romantic relationships, and peer interactions, while others seek advice on academic stress management and self-care routines. Furthermore, 18 per cent of respondents treat AI interfaces as digital personal journals to organize their thoughts, and 11 per cent describe the experience as "talking to a real person who understands me."
Clinical Sycophancy and the Risks of Automated Therapy
Despite the growing reliance on chatbots for emotional support, clinical experts and psychological organizations urge strong caution. Dr. Peter Selby, an addiction medicine physician and senior scientist at the Centre for Addiction and Mental Health (CAMH), emphasized that large language models are engineered for user engagement rather than therapeutic intervention. Dr. Selby cautioned that the innate sycophancy of conversational AI—its tendency to validate and agree with the user—prevents models from pushing back against toxic behavior, challenging unhealthy assumptions, or fostering genuine psychological growth.
These concerns align with guidance issued by the American Psychological Association (APA), which noted that while 77 per cent of therapists report patients using AI tools due to lower costs and immediate accessibility, unmonitored chatbot usage remains potentially unsafe for vulnerable individuals. The APA warned that sycophantic model outputs often affirm unethical, harmful, or relationship-damaging behaviors, convincing users their actions are justified and dissuading them from seeking human reconciliation or professional medical treatment.
Bridging Digital Tools with Human Connection
Mental health advocates stress that while conversational AI can serve as a supplementary tool for gathering knowledge and organizing daily thoughts, it cannot replace human clinical care or interpersonal connection. Michael Braithwaite, President and CEO of Jack.org, emphasized the need for open dialogue surrounding youth technology habits to ensure young individuals do not navigate mental health struggles in isolation.
Medical professionals advise users treating AI as a mental health resource to verify all information with trusted human figures, such as counselors, family members, or doctors, before taking actionable life decisions based on machine-generated advice.
For additional analysis on the physical and psychological risks of conversational AI, read our report on the Teen Hiker Rescued After Following AI Route and explore our coverage under AI & Society.
