Abdication

Abdication means the intentional design of a product feature in a way that ties a platform's hands on how it can moderate it - the platform Abdicates the majority of its capacity to do integrity work. The most common example of this feature is end-to-end encryption, which prevents platforms from having any visibility into the contents of the encrypted messages that they pass. While End-to-end encryption certainly has benefits for every user's privacy, and is enormously important for free expression within authoritarian regimes, it comes with the downside that platforms have no visibility into the content of messages, and thus cannot prevent the harm that those messages may cause. While end-to-end encryption is the clearest and most extreme example of abdication, there are other examples, like the use of blockchains for storage meaning that content can never be taken down.

Abdication stands apart from the other categories of preventing harm - it isn't a strategy for a platform to prevent harm, it's a strategy a platform employs to try to evade blame for the harm that it causes. While platforms that abdicate their content-moderation responsibilities may claim that they are powerless to act to prevent harmful content from circulating on their platforms, this is disingenuous. Instead, the platform simply made the earlier choice to abdicate this responsibility, and in doing so pre-baked in the ethical tradeoff they find acceptable. Platforms that abdicate responsibility understand (if not encourage) the creation and distribution of some harms (most notably, CSAM), while leaving the platform safer from some other forms of harm (most notably, limitations on free expression). When platforms abdicate responsibility in this way, we should be clear-eyed about, and hold them accountable for, the implications of the tradeoffs that they have chosen to bake into the design of their system.