Data Autonomy: Counter-Surveillance Strategies for Civil Society

Marlon Kautz, a grassroots organizer who is facing a prosecution tied to activism, opens with a history of social movements and the state responses to them. He traces the shift from informants and wiretaps in earlier decades to police work built on data platforms, where companies provide law enforcement portals for fast data requests. He argues that zero-knowledge messaging tools, often presented as the answer, have real limits: they face legal pressure and backdoor demands, and they are slow and costly to build and maintain. The organizations behind them can also become political targets.

The alternative he proposes is data autonomy, meaning a movement controls where its data lives, who has physical access to the machines that process it, and how decisions about that data are made. He argues this control should be collective and held by the political groups that depend on it, rather than by a single tech collective that ends up hosting everyone's data. The main obstacle is labor, so he suggests treating deployment and maintenance as code. He demonstrates Co-op Cloud, a set of recipes built on Docker Compose that deploys self-hosted applications such as Nextcloud from a command-line tool, while acknowledging that the work still requires skilled maintainers.

In the question period, audience members raised concerns that self-hosted servers could become identifiable honeypots and that technical power could concentrate in a few people. Kautz acknowledged these trade-offs, including the idea that spreading services across many servers can blur who is connecting to what. The takeaway for practitioners is to treat data infrastructure as core organizing work and to contribute to shared, maintained tooling rather than building one-off solutions.