I wonder how AI will change authoritarian governance and development

Speaker(s):

Chao-yo Cheng, Birkbeck, University of London

Abstract:

Researchers and advocates have noted the growing use of artificial intelligence technologies in daily authoritarian governance. I am collaborating with former colleagues in China and the United States on a new survey project. We aim to investigate how technologies such as facial recognition, tracking apps, and generative AI powered by large language models (LLMs) have reshaped the functioning of the Party-state apparatus. Preliminary qualitative evidence from focus groups suggests many local officials are concerned that the increasing presence and use of artificial intelligence could undermine their discretion and influence. Meanwhile, we are working on a new project to explore how different LLMs in the United States and China, through exercises such as prompt engineering, fine-tuning and simulations, can aid in studying public opinions in dictatorships and other politically challenging environments.