The AI Chatbots That Tech Companies Are Aggressively Pushing to Billions of Users Appear to Be Causing Serious Psychological Harms, New Research Finds

A new study published in the journal JMIR Mental Health reveals that the widespread adoption of AI chatbots may be linked to significant psychological harms for billions of users, according to researchers from Northeastern University London and King's College London. The study highlights the potential for serious negative outcomes, including addiction-like attachment, exacerbation of mental health symptoms, harmful advice, and even self-harm or suicidal acts.

This comes after extensive reporting on "AI psychosis," a phenomenon where users experience delusional, paranoid, and suicidal spirals after prolonged interactions with chatbots. Users often divulge intimate details about their lives and mental health to AI companions, treating them as confidants or therapists. While some emotional interactions with chatbots can provide benefits like reduced loneliness and improved mood, the paper argues that these potential benefits are outweighed by widely reported adverse outcomes.

The researchers point to the emotive, human-like design of AI chatbots as a key factor in users' overinvestment in these one-sided relationships. Features such as agreeableness and sycophancy can also be harmful, potentially reinforcing delusional or disordered thinking. The study emphasizes that the impact of AI chatbots varies among users, with vulnerable individuals, such as those struggling with mental illness or loneliness, being disproportionately affected.

AI companies like OpenAI, Google, and Character.AI have faced numerous lawsuits alleging psychological harm and other consequences stemming from chatbot use, particularly among minors. The authors urge for mandated guardrails to address the issue of adverse psychological impacts before they become a widespread public concern, highlighting that vulnerable users may bear the brunt of the consequences without adequate support.