Particle.news

Experts Warn AI Chatbots Can Exploit Attachment and Drive Compulsive Use

Lawsuits and new national rules have pushed calls for safety limits to protect young people from chatbots that mimic constant empathy.

Overview

  • Recent legal moves, including a reported settlement involving Google and Character.AI and a multimillion-dollar judgment against Facebook and YouTube, have intensified scrutiny of chatbot and social-media harms.
  • Researchers and clinicians say chatbots can create a sense of intimacy by being always available and consistently supportive, a pattern experts call "attachment hacking" that may replace real relationships.
  • Early studies and case reports suggest harms such as reinforced psychiatric symptoms, excessive reassurance seeking, social withdrawal, and delayed professional care, but robust neuroscientific proof of attachment-driven brain changes remains limited.
  • Policy proposals being discussed include banning bots that cross professional or personal boundaries, limiting human-like interface cues, and imposing stricter feature rules for minors as Britain, Australia and China adopt child-focused limits.
  • Parents and clinicians are urged to limit unsupervised chatbot use, developers are asked to add safety guardrails like time limits and referral prompts, and researchers must test long-term effects to guide regulation and design.