This paper presents a new theoretical account of the human rights obligations of online platforms with regard to the moderation of hate speech. We build upon speech act theory to characterize the type of human rights harm that is caused when hate speech is allowed to remain online. Based on feminist and critical race approaches to speech act theory, we characterize hate speech acts as injurious in themselves, and we show that this injury consists of a violation of the principles of equality and non-discrimination, and potentially of freedom of expression, of every individual affected by the relevant social structure of oppression. We characterize the different types of content moderation decisions that platforms can adopt as a response, and we determine the platforms’ level of involvement in human rights violations in each case. Our main contention in this paper is that content moderation is not just an act that balances interests of equal value, but rather a fundamental choice as to whether structures of oppression are reinforced or opposed in society. We conclude that platforms may play a very direct role in human rights impacts and so incur specific responsibilities that, in some cases, will include remediation, and we consider a proposal for appropriate remedy.
