价值冲突与治理出路:虚假信息治理中的人工智能技术研究

Value Conflict and Governance Solution: Examining the Role of Artificial Intelligence in Tackling Misinformation

人工智能技术不仅能将虚假信息的生产自动化,也能用于检测和甄别虚假信息。本文旨在探讨“用技术手段来解决技术隐患的快速方案是否有效以及如何使之有效”的问题。基于真实信息以及搜寻是有成本的这一前提,本文从信息搜寻的视角出发,解释人工智能在生产和核查虚假信息方面对用户信息搜寻成本以及搜寻环境的影响。尽管人工智能技术能够用于删除虚假账号和高效识别虚假信息,但这种技术性方案也存在包括提高用户搜寻负担,造成信息环境鸿沟等诸多长期风险。本文进而探讨了在“自动化”虚假信息泛滥的网络困境中,跨部门应如何在人工智能技术及算法的使用上进行规制与合作,以保障人们对于真实信息的搜寻。文章认为,在信息核查的实践中不能忽视人工智能的公共性特征,一个连结各相关方的第三方机构应被建构,专业事实核查机构和新闻记者也应被给予更多支持。

Artificial Intelligence(AI) has learnt to generate fake media creations overnight, while AI-based tools could also be used to debunk fake news. Considering the rampant fake accounts and social bots on social websites, individual uses have to pay costs to achieve the real stories. We lay out Stigler’s model of search as a framework and illustrate how AI influences the information market and the environment that encouraging real news. Although more AI can help to combat fake news and disinformation more effectively, the burden on individuals for searching real news would not be reduced and the disparity in the information environments among various populations increases in a long run. We then develop the strategies and a cross-agency cooperation model for AI use in fighting fake news and disinformation. We highly recommend that when dealing with publicinterest-related news, AI algorithm should be transparent and the institutions should be held responsible for algorithms. Practically, a third-party organization that is independent from both the industry and government should be established, and also, the fact-checking organizations and professional journalists should be given more support.