Democracy is a system: a complex nonlinear tightly coupled system of systems that's surprisingly vulnerable to hacking, subversion and exploits of all kinds. Our community has expertise in securing complex systems. Can we apply that thinking to design of democratic systems, especially socio-technical systems in the age of increasingly pervasive data collection, ubiquitous computing, and capable AI?