Google study: anthropomorphic chatbot language backfires with users

A Google Research study to appear at AIES 2026, using LLM-based social simulation across the US, UK, and Indonesia, finds that anthropomorphic companionate language from chatbots lowers perceived likability, humanness, and trust—especially among women and older users.

2026-09-03 ~ 2026-09-03 · 4 related posts

• Google Study: Chatbots Using Companion Language Like "I'm Worried About You" Feel Less Trustworthy — jacyanthis · 2026-09-03
• Google Researchers Use LLM Social Simulation to Test 16 Chatbot Behaviors Across 4 Countries — jacyanthis · 2026-09-03
• Google Study: Women and Older Users Rate Anthropomorphic Companion Chatbots Lower, Effects Vary by Country — jacyanthis · 2026-09-03

1 near-duplicate retellings: jacyanthis