Governing Artificial Intelligence: Public Preferences and Regulatory Options
创建于 更新于
摘要
This study investigates how citizens across seven countries evaluate competing AI governance designs using a three-dimensional conjoint experiment (objective: safety vs. innovation; mode: public vs. private; level: national vs. international) across three domains (workplace, policing, warfare). It finds broad support for AI regulation with a robust preference for safety over innovation, public governance over private self-regulation, and international over national regulation. The strength of the safety preference is higher among those who perceive AI as risky, unpredictable, and personally consequential. The findings imply a systematic misalignment between dominant national models (e.g., US, China) and citizen preferences, highlighting the political salience of risk perceptions in AI governance [page::0].
