Being Watched: Embedding Ethics in Public Cameras

‘Smart cities’ infrastructures have won fans among many officials, city planners, and infrastructure manufacturers. A collection of hardware devices such as cameras, sensors, and controllers linked to software-driven intelligence—often with an Internet, grid or network element and deployed at scale—these systems promise cities, regions and even nations better services, improved management capabilities and sometimes lower costs. Increasingly, the voluminous and diverse data they produce will be analyzed by AI or machine-learning routines, raising questions regarding how those machine-based procedures operate and the ethical systems undergirding their analyses.  

As more of such city-based infrastructures move toward automated and algorithmically-optimized capabilities, the human and broader machine and platform environments present legal, ethical and accountability issues. These affect the strata of planners and managers in cities, the technologists producing (and selling) the software and hardware systems and companies, as well as citizens living and working with these systems.  

First, we propose to design a system of differential access to the data produced by the camera systems used by municipalities as a first step toward maximizing data usability, minimizing privacy concerns, and working toward increased trust for these systems. Computer vision research undergirds this technological challenge. Second, the design and build approaches will rely on insights from research investigating aspects of public trust in technology systems, and in city governance more broadly. Intensive fieldwork will use focus groups, surveys, and public meetings for gathering data that can identify the sources of threats as well as benefits, and how these are framed. Third, we will develop and test public engagement best practices so that cities and their citizens reciprocally shape the best policies around ‘smart cities’ technologies. A core component considers cities’ data sources and data policies. Finally, we will analyze state open records laws to better understand the legal requirements for collecting, storing and maintaining the video/audio data produced by smart cities technology. Building off this analysis, we will determine best practices for smart cities data as well as drafting model legislation that state legislatures can use to update their public records laws in accordance with best practices. 

Our approach relies on concepts from science and technology studies to explore how we can design and build better human and machine systems to insure these monitoring and management systems perform fairly, accountably, and in ways that optimize trust from the people producing and using the data.   

Team Members
Atlas Wang
Electrical and Computer Engineering
Amy Sanders
School of Journalism and Media
Select Publications

Sharon Strover, Maria Esteva, Tiancheng Cao, and Soyoung Park. “Public Policy Meets Public Surveillance.” AoIR Selected Papers of Internet Research.