Deconstructing the ethics of large language models from long-standing issues to new-emerging dilemmas: a survey
Aug 13, 2025·,,,,,,,,
,,,,,,,,·
1 min read
Chengyuan Deng
Yiqun Duan
Xin Jin
Heng Chang
Yijun Tian
Han Liu
Yichen Wang
Kuofeng Gao
Henry Peng Zou
Yiqiao Jin
Yijia Xiao
Shenghao Wu
Zongxing Xie
Weimin Lyu
Sihong He
Lu Cheng
Haohan Wang
Jun Zhuang
Abstract
This survey deconstructs the ethics of large language models, mapping long-standing issues such as bias and misinformation to newly emerging dilemmas including agentic behavior, environmental cost, and cross-cultural alignment. We outline a research agenda for ethics-aware LLM development.
Type
Publication
AI and Ethics, Volume 5, pages 4745–4771 (2025)
Abstract
This survey deconstructs the ethics of large language models, mapping long-standing issues such as bias and misinformation to newly emerging dilemmas including agentic behavior, environmental cost, and cross-cultural alignment.
