Weaponizing Censorship Infrastructure

Abstract

Nation-states that engage in Internet censorship pose well-known threats to the openness of communication to those within their borders. In this talk, we will show that censorship infrastructures can pose significant threats to the entire Internet. We have discovered that malicious actors can weaponize censorship infrastructure to launch attacks. We will briefly describe two attacks in particular: (1) How to trick censors into blocking a victim’s connections, and (2) How to leverage censors to launch high-volume denial of service (DoS) attacks against arbitrary victims—and how we used AI to discover them. Collectively, these results will show that the negative impact of censorship extends well beyond the censor’s borders, and that they pose an even larger threat to the Internet as a whole than previously understood.