Throughout the U.S., cities have addressed smart city tech policy by developing smart city legislation, some passing ordinances tied to specific surveillance technologies, while others are developing road maps outlining a vision and guiding principles. Sunlight’s Open Cities team  identified a set of underlying principles from these policies that smart cities should prioritize in order to ensure adequate public accountability and transparency as they adopt new technologies.

The principles identified here sample the most consistent and effective ways we’ve seen cities promote transparency, public disclosure, and accountability with smart city technology. While few cities embodied all the principles in one document, we found that they are in fact modular, with cities placing emphasis on different principles depending on their local context.

Even so, we think cities should strive to implement all the principles, as transparency provides the public with the ability to ensure that government is accountable in the protection of rights, and gives city residents an opportunity to advocate for their own interests.

Smart Cities Best Practices

  1. Engage early and often
  2. Make decisions in the open
  3. Establish smart governance
  4. Document public privacy and impact assessments
  5. Develop robust open data practice

Engage early and often

Public officials need to engage community members in determining what technologies to acquire when they can potentially affects issues of privacy and security. This includes publicly deciding how personal data is collected, used, and protected on behalf of the public good. Moreover, community members should be a part of deciding if the benefits outweigh the costs of new smart city projects, especially when considering the non-monetary costs associated with surveillance (such as loss of privacy and civil liberties). This can be done by holding public hearings, community meetings, and creating online forums to inform residents about potential projects, solicit their opinions, and address their concerns.

Smart city policies should include concrete commitments to community engagement throughout every phase of their initiatives or tech rollouts, for example committing to hiring staff trained in community engagement strategies, committing to building teams with diverse backgrounds, and committing to maximum opportunities for feedback loops before, during, and after smart city projects. There should be a commitment to soliciting feedback from diverse stakeholders, including university partners and the local tech community for subject matter expertise and education;

Open decision-making

To build trust and ensure acquired technologies are developed in ways that do not violate civil liberties, smart city policies should mandate that public officials must share how and why smart city projects are selected before their official implementation. To this end, there should be a public discussion of plans for new smart technology acquisition even before the procurement process begins. This includes explaining what the problem is, why this smart technology tool offers a solution, what alternative solutions exist, and why this is the best solution.

This can be facilitated through the creation of a priority framework or rubric that outlines how projects are selected. This information should be publicly available, easily accessible, and subject to public comment.

Smart cities policies should hold public officials to clear transparency and public accountability standards around technology implementation. This means legislating that staff responsible for overseeing smart city policy are on the hook to disclosing how, where, and why a project will be implemented within a specific timeline of a new smart city project being launched. Wherever possible, the relevant plans and documents should be published throughout the project selection and implementation process. This includes, but is not limited to, publishing a Smart City Strategic Roadmap outlining upcoming smart city project objectives and priorities and/or issuing a notice of anticipated projects (prior to the Requests For Proposals) to facilitate conversation. Solicitation documents, RFPs, budgets, and implementation timelines should be published in a timely manner. During the implementation process, if possible, publish locations of sensors and smart infrastructure. Where this is not possible, at a minimum the city should expect to document and publish how implementation locations are selected in a timely manner.

These methods are crucial in determining if a technology will disproportionately affect a subset of the population and in developing strategies to minimize harm. Cities in the United States cannot ignore the fact that historically, policies around data collection and surveillance have disproportionately affected already vulnerable and marginalized communities. By ensuring mechanisms for public accountability around the deployment of new technologies, public officials allow for both residents and experts to provide oversight and minimize adverse effects.

Good smart city governance

Smart city policies should set an expectation of public disclosure of the types of data being collected, as well as methods of data storage and transfer. This can be facilitated through the creation of a data inventory, to record basic information (including, but not limited to: name, contents, update frequency, use license, owner/maintainer, privacy considerations, data source) about data collected from smart technologies

In addition, public officials should thoroughly consider data ownership, especially when entering into a public-private partnerships and should ensure cities have the ability to own or control the data generated from the devices and platforms they are using.  There should be clear guidelines on which parties data can be shared with, and in what situations data sharing will be allowed. These guidelines should be codified in third-party vendor contracts. (GovEx provides template language within their Data Ownership and Usage terms for Government Contracts).

Cities should also maintain transparency around data analysis and usage. If data is being treated with algorithms, mechanisms used to automate processes and data analysis, these algorithms should be publicly available and open to public and expert oversight. Wherever possible, the algorithms should be subject to an audit to ensure the algorithm is being used as intended. While algorithms are upheld as methods to increase objectivity, the design of an algorithm can be subject to bias (the bias of the algorithm creator as well as any bias in the data collected).

Documented consideration of privacy implications

Smart city policies should hold public officials to consider how new technologies will impact a resident’s right to privacy and ensure there are appropriate safeguards to protecting this right. Proactive community engagement is crucial in this instance, because the definition of what the right to privacy entails should be jointly defined by public officials and community residents. Officials should also record and publicly disclose these considerations, to ensure maximum accountability and build trust that the city took all the necessary measures to protect residents.

A smart city policy should outline how a city will consider the privacy implications for each smart technology project. This can take the form of a privacy impact assessment, created prior to each smart technology acquisition. To allow for effective public oversight, a privacy impact assessment should be completed, publicly available, and subject to a public review period before the technology is acquired. At a minimum, officials should clearly define how Personally Identifiable Information will be handled and protected with regards to smart technology deployment.

Robust open data practice

Cities that are striving to increase data collection and real-time analysis for improved efficiencies recognize that data transparency is integral to a true smart city. As a result, smart cities should  default to open data and treat information as a public good. Cities that prioritize open data through open data portals and open data infrastructure allow for more creative uses of data and provide new opportunities for public engagement and citizen empowerment. A vision of an open, smart city will require prioritizing digital inclusion and tackling the digital divide.

A policy creating an open data practice creates a mechanism to publish data (not personally identifiable information) collected from sensor technology. For tools on building a robust open data practice, see Sunlight’s Open Data Policy Guidelines.

A Smart Policy for a Smart City

Portland’s Framework, SmartCity PDX begins with a quote from their mayor, Mayor Ted Wheeler: “As a city, we cannot continue to let technology happen to us.” As cities strive towards a smart, Jetsons-like future, they must be thoughtful about how they bring on new technologies. When a city defines a roadmap or creates guidelines to how it plans to become a smart city, it begins the process of creating a unified vision for how it will incorporate technology into city planning and how it will center users (in this case, the public) in the design. A city will not become smart simply by increasing data collection and prioritizing data-driven decisions. A truly smart city will increase its capacity to be effective in providing for residents, while addressing the legitimate concerns that they may have with these technologies.

Thank you to all the individuals who took the time out to speak with me about their work with smart cities, smart technologies, surveillance technologies, AI that made the collection of these resources possible.

Danielle DuMerer, Chief Innovation Officer at City of Chicago; Charlie Catlett, Senior Computer Scientist at Argonne National Laboratory; Heather Patterson, Senior Research Scientist at Intel; Mayoral Appointee to Oakland Privacy Advisory Commission; Brian Hofer, Oakland Privacy Advisory Commission Board Member; Christine Kendrick, Smart Cities Coordinator at City of Portland (OR); Hector Dominguez Aguirre, Open Data Coordinator at City of Portland (OR); Sari Ladin-Sienne, Chief Data Officer at City of Los Angeles; Kiana Taheri, Smart City Coordinator at City of Los Angeles; Rashida Richardson, Director of Policy Research at AI Now; Peter Rowland, Research Fellow at UC Berkeley; Steve Trush, Research Fellow at UC Berkeley; Tara Pham, Co-Founder and CEO at Numina; Sawyer Middeleer, Research Assistant at Metropolitan Planning Council; Ellen Hwang, Assistant Director of Strategic Initiative at City of Philadelphia