Social-media companies make extensive use of artificial intelligence in their efforts to remove and block terrorist content from their platforms. This paper begins by arguing that, since such efforts amount to an attempt to channel human conduct, they should be regarded as a form of regulation that is subject to rule-of-law principles. The paper then discusses three sets of rule-of-law issues. The first set concerns enforceability. Here, the paper highlights the displacement effects that have resulted from the automated removal and blocking of terrorist content and argues that regard must be had to the whole social-media ecology, as well as to jihadist groups other than the so-called Islamic State and other forms of violent extremism. Since rulebylaw is only a necessary, and not a sufficient, condition for compliance with rule-of-law values, the paper then goes on to examine two further sets of issues: the clarity with which social-media companies define terrorist content and the adequacy of the processes by which a user may appeal against an account suspension or the blocking or removal of content. The paper concludes by identifying a range of research questions that emerge from the discussion and that together form a promising and timely research agenda to which legal scholarship has much to contribute.

中文翻译：

---

监管社交媒体上的恐怖主义内容：自动化和法治

社交媒体公司广泛使用人工智能来从其平台上删除和阻止恐怖主义内容。本文首先认为，由于此类努力相当于引导人类行为的尝试，因此应将其视为一种受法治原则约束的监管形式。然后，本文讨论了三组法治问题。第一组涉及可执行性。在这里，论文强调了自动删除和阻止恐怖主义内容所造成的流离失所效应，并认为必须考虑到整个社交媒体生态，以及所谓的伊斯兰国和圣战组织以外的圣战组织。其他形式的暴力极端主义。由于统治经过法律只是遵守规则的必要条件，而不是充分条件——的-法律价值观，然后本文进一步研究了两组问题：社交媒体公司定义恐怖主义内容的清晰度以及用户可以针对帐户暂停或阻止或删除内容。本文最后确定了讨论中出现的一系列研究问题，这些问题共同构成了一个有前途和及时的研究议程，法律学术对此有很大贡献。