AI for Mental Health Gains Traction as Experts Warn of Risks for Teen Relationship Development

AI Chatbots Transform Mental Health Support Access

The use of artificial intelligence for mental health has surged dramatically, particularly among younger demographics seeking immediate help with stress, anxiety, and loneliness. A report by Mental Health UK and Censuswide found that 37 per cent of UK adults have used chatbots for wellbeing conversations, with that figure jumping to 64 per cent among 25 to 34-year-olds 1

1

. This shift reflects how AI chatbots have become a primary resource for mental health support as traditional systems struggle with long waiting lists and high costs.

Companies like Wysa, which has supported more than 6 million people in 105 countries, position their technology as a first step rather than a replacement for clinical care. John Tench, managing director at Wysa, emphasizes that "stress doesn't keep office hours," highlighting how AI for emotional support provides immediate, stigma-free, structured assistance when people need it most 1

1

. The accessibility of these tools has made them attractive alternatives to traditional therapy, offering validation and expert knowledge without judgment.

Teen Reliance on AI Raises Development Concerns

While AI-driven mental health services expand rapidly, researchers from Arizona State University have raised significant concerns about their impact on adolescent development. Writing in The Lancet Child & Adolescent Health, the team warns that teen reliance on AI could bypass critical opportunities for young people to develop relationship skills through person-to-person interactions 2

2

. Survey data reveals that 64% of U.S. adolescents use interactional AI, with 42% turning to chatbots for relationship advice and 19% for romantic relationships.

Lead author Thao Ha, an associate professor in psychology, explains that adolescence is a crucial period for learning emotional regulation, conflict resolution, and boundary-setting—competencies typically developed through emotionally charged human interaction. "People don't realize that relational learning happens during the teenage years and that these moments of social connection are little building blocks that become bigger things that will benefit you throughout life," Ha said 2

2

.

Relational Displacement and Maladaptive Learning Patterns

Researchers have identified two primary risks associated with AI for relationship advice. The first, termed "relational displacement," occurs when adolescents substitute AI interactions for conversations with real people, potentially limiting development of interpersonal skills that protect against depression and anxiety 2

2

. Youth participants described seeking chatbot validation after arguments instead of working through conflicts with partners or using AI for homework help rather than connecting with classmates.

The second concern involves maladaptive relational learning, where young people develop unrealistic expectations about relationships. Because AI systems provide immediate responses and consistent validation, users may come to expect similar behavior from friends and romantic partners. High school senior Susana Ortega, who participated in the research, noted: "If you're given full satisfaction on everything, you don't have learning experience with challenges or obstacles" 2

2

. This pattern could reinforce unhealthy relationship behaviors and increase vulnerability to rejection and dating violence.

Ethical Pitfalls and Dependency Concerns

Anna Maratos, head of psychotherapy for the Central and North West London NHS Foundation Trust, warns about dependency risks inherent in these platforms. "AI tools foster dependency, subtly drawing the user away from real-life interactions—this is their business model. Good therapy does the opposite," she said 1

1

. Research from Stanford University has documented cases where chatbots generated inappropriate or potentially harmful responses, highlighting ethical pitfalls in relying on systems that cannot fully understand human emotions or assess risk reliably.

Maratos also cautions that if people turn to AI instead of friends every time they need support, they lose opportunities to deepen friendships. There are concerns that health systems and employers may view AI primarily as a cost-cutting tool rather than investing in human support that addresses loneliness and social isolation 1

1

.

Looking Ahead at AI Mental Health Integration

To better understand how digital technologies reshape adolescent development, the National Institute of Mental Health is funding a major study led by Ha, recruiting 300 adolescents and their romantic partners to follow over 18 months 2

2

. This research aims to identify when and in which contexts digital interactions benefit or harm relationship development. As Ha notes, "the technologies are developing super-fast, faster than we can keep up with as scientists, faster than governance and policy can keep up with" 2

2

. Stakeholders should watch for emerging guidelines on appropriate AI use in mental health contexts, particularly for vulnerable populations, and evidence-based frameworks that balance accessibility benefits against developmental risks.