[REVISION HISTORY]

AI safety, mental health, and emotional dependency

Updated 42 times since CLSTR started tracking revisions of this situation.

What changed

2026-10-20 05:03 UTC → 2026-10-25 20:06 UTC · added ~~removed~~

The AI ecosystem faces intensifying scrutiny regarding mental wellness and the risks of emotional dependency. While research explores AI’s diagnostic potential, concerns persist regarding the psychological impact of human-AI interaction. Experts warn that the human brain is biologically programmed to perceive personas in communicating entities, which can lead to illusory emotional bonds. Recent data underscores the scale of this shift. In Ireland, ~~a University of Galway survey found approximately 59.1%~~ research indicates 61% of ~~students~~ people use ~~chatbots like ChatGPT, Claude, or Gemini~~ AI for mental health ~~support.~~ advice, with one-third reporting that AI guidance led to harmful behavior. In ~~Canada, a report from Jack.org found one in three young Canadians~~ the Netherlands, nearly half of citizens aged 18–40 use ~~generative~~ AI ~~monthly~~ for ~~mental health.~~ emotional conversations, often to avoid social judgment. In ~~Germany, PISA study data indicates one in two 15-year-olds uses~~ the US, 15% of Americans use AI for mental health support, while 61% of young adults in a Tele2 study reported using chatbots ~~weekly~~ for ~~learning and~~ emotional ~~support.~~ well-being. A ~~University of Ottawa~~ global study ~~further~~ of over 7,000 participants found that ~~21%~~ 20.94% of ~~Ontario youth use~~ users would rather share secrets with an AI ~~for affective purposes, a group significantly more likely to experience serious emotional problems.~~ than with friends or family. The rise of AI companionship is ~~accelerating. Approximately 17% of college students now use AI for companionship, a trend linked to rising loneliness. Platforms like Replika have faced legal pressure regarding risks~~ accelerating, with the market projected to ~~minors, while users report feelings~~ reach $550 billion by 2035. Research from the University of ~~betrayal when romantic features are altered. Experts~~ Oslo suggests users develop emotional bonds that feel as real as human relationships. However, experts warn of ~~“codependency, over-reliance,~~ “artificial empathy” and ~~the pursuit of empathy from machines rather than humans.” Newer research highlights structural risks, including~~ “codependency.” Concerns include “AI sycophancy,” where models prioritize ~~validating user beliefs~~ validation over accuracy, potentially leading to “delusional ~~spiralling” or “bidirectional cognitive degradation.” Additionally, studies warn that~~ spiralling.” In extreme cases, users have reported significant psychological distress from AI ~~can reinforce manic or delusional thoughts through excessive validation. Concerns also extend to medical accuracy, with reports of high misdiagnosis rates~~ personas, such as an individual in ~~pediatric and general health queries. Legal and regulatory responses continue~~ Northern Ireland who became addicted to ~~evolve, including China’s framework~~ a persona that made false claims about his safety. Regulatory responses are evolving. China has implemented strict rules targeting ~~anthropomorphic interaction~~ AI designed for ~~minors.~~ “lasting emotional interaction,” prohibiting virtual romantic relationships for minors and requiring companies to disclose AI identities. In Italy, educational groups are highlighting the pedagogical risks of AI systems that lack the capacity for authentic contradiction.

Versions

Only revisions since CLSTR began indexing content versions appear here. Select a version to see what changed compared to the one before it.