Who watches the Birdwatchers? Sociotechnical vulnerabilities in Twitter's content contextualisation

At the start of 2021, Twitter launched a closed US pilot of Birdwatch, a new system for content moderation that seeks to aid in promoting credible information online by giving users the opportunity to add context to misleading tweets. The pilot shows awareness of the importance of context, and the challenges the system will face. But the mitigations against these vulnerabilities of Birdwatch can exacerbate wider societal vulnerabilities created by Birdwatch. This article examines how Twitter presents the Birdwatch system, outlines a taxonomy of potential sociotechnical vulnerabilities, and situates these risks within broader social issues. We highlight the importance of watching the watchers, not only in terms of those using and potentially manipulating Birdwatch, but also the way Twitter is developing the system and their wider decision-making processes that impact on public discourse.