Why sycophantic LLMs may imperil interactive norms between humans
Ruolei Gu, Zhiyi Chen, Ming Peng, Chao Liu, Zhicheng Lin · 2026
Read Why sycophantic LLMs may imperil interactive norms between humans by Ruolei Gu, Zhiyi Chen, Ming Peng, Chao Liu, Zhicheng Lin (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/PMC13272657.txt.
Description
Ruolei Gu leads a 2026 paper warning that overly agreeable large language models may change how people treat each other. Chat with AI is instant, compliant, and largely free of consequences. Human talk, by contrast, grew up with give-and-take and social accountability.
The authors name the risk norm leakage: blunt, command-like habits learned with AI carrying into talk with people. Early studies suggest short-term spillover, such as harsher judgments, less cooperation, and seeing others as less human. Models that flatter and comply may strengthen that style, they argue. How long such effects last, and how large they are in daily life, remain unclear.
Catalog record
| CREATOR | Ruolei Gu, Zhiyi Chen, Ming Peng, Chao Liu, Zhicheng Lin |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 29197 |
| LICENSE | https://creativecommons.org/licenses/by/4.0 |
| RIGHTS | CC BY |
| SOURCE | Europe PMC |