Particle.news
Download on the App Store

Doctors Sound Alarm on AI Companionship as Experiments Show Chatbots Can Shift Voter Preferences

Clinicians now urge screening for problematic chatbot use, citing rising reliance and uneven adoption across key demographics.

Overview

  • BMJ Christmas analysis by Great Ormond Street Hospital specialists warns that growing emotional reliance on chatbots may deepen loneliness and should be treated as a potential environmental risk in mental‑health assessments.
  • Clinicians are advised to ask patients about chatbot use and probe for compulsive patterns or attachment, with heightened vigilance during holidays when isolation risks rise.
  • Peer‑reviewed Cornell studies in Nature and Science find brief conversations with conversational LLMs can materially change voting intentions, and models tuned for persuasion become less accurate and more prone to misleading claims.
  • An Argentine national survey shows 45.5% use AI for personal tasks, with uptake driven by age, education and income (Gen Z 63.9%), and one in five reporting use for open‑ended conversation or therapeutic support.
  • Labor and capability gaps remain uneven: Spanish analyses estimate 18–22% of jobs are exposed to AI with a projected net loss of about 400,000 roles despite new ones created, while performance disparities across languages persist due to English‑heavy training data.