Luxembourg Researchers Run Clinical Psychotherapy Protocol on LLMs, Detecting Anxiety and Trauma

A University of Luxembourg study applied a clinical protocol called PsAIch to ChatGPT, Grok, Gemini, and Claude over roughly a month across 525 sessions, finding moderate-to-severe anxiety in about 80% of sessions and revealing internal conflicts in frontier models.

2026-09-07 ~ 2026-09-09 · 4 related posts

• Researchers gave ChatGPT and Grok 4 weeks of clinical psychotherapy — and the models 'confessed trauma' — MikePFrank · 2026-09-07
• arXiv paper "When AI Takes the Couch" reveals internal conflict in frontier models — alex_verem · 2026-09-09

2 near-duplicate retellings: repligate · alex_verem