Regulating Online Content Moderation

Regulating Online Content Moderation
复制标题

规范在线内容审核

DOI:
--
复制
发表时间:
2017
期刊:
影响因子:
--
通讯作者:
Kyle Langvardt
Kyle Langvardt
中科院分区:
--
文献类型:
--
作者:
Kyle Langvardt

文献摘要

被引文献

相似文献

最高法院在 2017 年裁定,“互联网上的广大民主论坛,特别是社交媒体”是“……交换意见的最重要场所”。然而在这些论坛中,发言者受到了世界上最严格、最迅速的审查制度。这种审查制度不是来自政府,而是来自少数私营公司——Facebook、Twitter、谷歌——以及大量的人工和算法内容审核者。内容版主的工作不可或缺;如果没有它,社交媒体用户就会淹没在垃圾邮件和令人不安的图像中。与此同时,内容审核实践仅与第一修正案的价值观松散对应。最近泄露的 Facebook 内部培训手册显示,其内容审核做法仓促、临时,有时甚至语无伦次。 现在是考虑立法保障网络空间中有意义的言论权的时候了。本文评估了解决该问题的一系列可能的方法。其中包括 1) 行政监控和合规制度,以确保内容审核政策接近第一修正案原则; 2)“个人责任”制度,将内容审核的控制权移交给用户; 3)相对简单的要求,要求公司披露其审核政策。每一种方法都存在严重的陷阱,但没有一种方法比选项 4) 更危险:继续将网络言论权委托给私营部门。
The Supreme Court held in 2017 that “the vast democratic forums of the Internet in general, and social media in particular,” are “the most important places…for the exchange of views.” Yet within these forums, speakers are subject to the closest and swiftest regime of censorship the world has ever known. This censorship comes not from the government, but from a small number of private corporations – Facebook, Twitter, Google – and a vast corps of human and algorithmic content moderators. The content moderators’ work is indispensable; without it, social media users would drown in spam and disturbing imagery. At the same time, content moderation practices correspond only loosely to First Amendment values. Recently-leaked internal training manuals from Facebook reveal that its content moderation practices are rushed, ad-hoc, and at times incoherent. The time has come to consider legislation that would guarantee meaningful speech rights in online spaces. This Article evaluates a range of possible approaches to the problem. These include 1) an administrative monitoring and compliance regime to ensure that content moderation policies hew close to First Amendment principles; 2) a “personal accountability” regime handing control over content moderation to users; and 3) a relatively simple requirement that companies disclose their moderation policies. Each carries serious pitfalls, but none is as dangerous as option 4): continuing to entrust online speech rights to the private sector.