My doctoral work examines what democratic societies lose and gain as our civic information infrastructures becomes increasingly technology- and AI-driven. This thesis documents three tech-driven changes to civic information ecosystems: the increased ubiquity of platforms, the growth of localized social networks, and the emerging influence of generative AI in local settings. With two studies conducted during the Covid-19 pandemic, I demonstrate how the platformization of news splintered the local information ecosystem in a high-stakes crisis environment. I then describe the rise of localized social networks, and conduct empirical studies that identify and measure the potential democratic benefits to communities of user-generated place-based content despite quality and bias concerns. Finally, I showcase how local generative AI summaries on search engines may amplify these biases in local data quality, and frequently lean on non-local sources. Through large-scale surveys, content analyses, experiments, and qualitative interviews, I arrive at my final argument that we should move towards "glocal" civic information ecosystems, where automated place-based monitoring is augmented by community-embedded investigations for high-stakes, crisis, or nuanced topics. Technologies can spread reliable civic information further and faster, but cannot make up for a lack of high-quality information or care.