The notion of a “smart” city has too-often been reduced to the use of technology to automate processes toward more efficiency—and frequently beginning with transportation and approaches that effectively equate city dwellers packages and materials that need to be efficiently moved through and processed in a factory.  Deeper challenges to cities and their governance systems are not automation challenges and require more than technology as they are symptoms of underlying challenges related to opportunity, fairness, perceptions, and trust.  In this sense, technology has in fact introduced new challenges for cities—and for democracy.  Continual improvements in miniaturization, power, and cost of computing devices, combined with similar improvements in digital imaging and artificial intelligence (AI) algorithms, provide the foundation for what we have termed “software-defined sensing.”  That is, a camera (and/or microphone) and a co-located computer (”edge computing,” or a computer at the “edge” rather than center of the network) can be installed and subsequently programmed to harness computing power (inexpensive, ever-expanding) to observe, or sense, phenomena and conditions that previously required a human observer (expensive, finite).  Software-defined sensing builds on prior improvements in AI on powerful servers that supported a transition in CCTV systems from expensive, human retrieval and analysis (e.g., to provide more detail about an event of interest) to automated, continuous analysis (to discover events of interest).  Whereas democracies are built on principles such as “innocent until proven guilty” and prohibitions on “unreasonable search” (that is, without probable cause) by government authorities, technology is now available to allow a governing authority to continuously observe the location (through smartphones) and activities (in surveilled public places) of individuals--introducing a crisis in democracy.  We will discuss the experience of the Array of Things project in Chicago, which has deployed hundreds of software-defined sensing systems in partnership with both the governing authorities and city dwellers. In particular, we will examine and evaluate the policies (established in 2016) designed to simultaneously harness the power of software-defined sensing and mitigate the associated privacy risks of Chicago residents.