Chat apps such as WhatsApp, Telegram, and Signal are increasingly popular platforms for communication. Their sometimes-closed nature and encryption affordances present researchers, governments, and law enforcement with unique problems of access, traceability, and, ultimately, understanding. It also makes them useful vectors for sowing disinformation. This research assumes a multi-platform perspective, describing the particularities of how chat apps can be used toward disseminating mis- and disinformation by way of cascade logic—the means by which information in chat app ecologies is trafficked upstream (making its way from private conversations into the mainstream) as well as downstream (allowing information to withdraw from the public eye), providing space for distortion along the way. Cascade logic also describes how chat apps allow the gradual withdrawal and self-segregation of individuals into, or emergence out of, layered spaces of privacy and obfuscation. We present an interview-based study exploring chat apps in three countries, synthesizing unifying dimensions across cultures and contexts: India, the United States, and Mexico. We analyze data from in-depth conversations with 33 individuals who work to either produce or track political content on chat apps. These interviewees work for a wide array of organizations: political parties, governments, extremist groups, digital political consultancies, news entities, and civil society organizations. We reveal key insights into the tactics of producers of political content on chat apps and show how these platforms are particularly suitable for harnessing human connections, or leveraging communities of trust, to sow disinformation.