New research shows that police forces across Canada are building extensive digital surveillance hubs without any public engagement. Smart city projects use very similar technologies with the same dangers, yet here residents and municipalities are increasingly implementing Open Smart City principles to avoid potential harms and strengthen public oversight. The police should not be exempt from democratic accountability and the same principles can be applied to them to rebuild it.
People around the world are recognizing the potential of emerging “smart” technologies—those technologies that use machine learning, artificial intelligence and large-scale data analysis—to provide more efficient and effective services. However, there is also significant potential for them to cause harm around privacy, discrimination, transparency, and the corporate capture of what are publicly and democratically controlled tools of government.
This danger is increasingly well recognized by both the public and governments, and in 2017 the Canadian federal government conducted the Smart Cities Challenge, a $75 million initiative that demonstrated a clear alternative approach to the failed attempt by Google’s Sidewalk Labs to launch a smart city project in Toronto. Among many elements, the Smart Cities Challenge winners stood out in contrast to the Sidewalk Lab’s project through their foundational emphasis on resident-led design and development.
The importance of having deeply transparent and democratic principles underpinning the development and implementation of potentially transformational but also potentially dangerous smart technologies is clearly well understood by both the general population and the government. Indeed, the Government of Ontario has pledged to continue to develop its artificial intelligence framework based on Open Government Partnership principles, a set of values that cement transparency, privacy, harm reduction and public engagement into the process.
Yet, how is it that the same kinds of data collection and analysis technologies, financed by the same municipal and provincial bodies, are exempt from these democratic principles when used by police forces?
Extensive research, including a recently completed 4-year doctoral investigation into the rise of smart technologies in Canadian policing by one of the authors, shows that, over the last two decades, Canadian police forces have replaced their emphasis on a strategy of ‘community-based policing’ with one of ‘intelligence-led policing.’ This strategy eschews building relationships of trust with communities to understand what they are experiencing and instead deploys a growing digital surveillance machinery to extract a bird’s eye version of that information without consent or oversight.
At the core of this strategy is the building of “real-time operations centres” (RTOCs) in police services across Canada: high-tech surveillance hubs modelled directly after U.S. Fusion Centres, deeply controversial mass surveillance units built post-9/11 for domestic counterterrorism programs.
These RTOCs bring together the same kinds of smart, AI-driven, surveillance-based technologies that many smart city projects utilize, and indeed they frequently integrate existing smart city systems like data from public transit cards1 or private and public CCTV networks into their surveillance apparatus.
The fact that there is not at least the same level of public oversight into and control over the police procurement and usage of these technologies poses a serious democratic deficit.
RTOCs in Canada
While appearing under a variety of names, RTOCs are already operational in almost all major municipal police services across Canada.
Based on thousands of pages of access to information requests, on-site visits and interviews, Thomas Linder2 pieced together how over the last decade such centres in Niagara, York Region, Ottawa, Calgary, Edmonton, Vancouver and more have become the go-to solution for Canadian police forces looking to “modernize” for the “digital age.”
Much like cities looking to smart city technologies, these police services are responding to a perceived need to operate effectively in an increasingly digital society. However, this research shows that police forces have avoided a public debate about the best approach and, instead, unilaterally adopted a mass surveillance model developed by the US military and domestic counterterrorism agencies in the aftermath of 9/11.
This model involves centralizing and expanding surveillance capabilities to reach far beyond those parts of society that police previously had access to, and doing so with tools that have a well-established potential for abuse and discrimination.3 These tools are developed by corporations like IBM, Palantir and Motorola Solutions who build similar technologies for U.S. and Canadian military and national security agencies and are also frequently contracted to help guide the police services in developing their RTOCs. The militarized counterterrorism model is baked into these Centres from the start.
While police have long used CCTV in a few key areas, RTOCs enable the deployment of many more cameras by linking existing private (like campuses and malls) and public (like parks, transit, or roads) networks.
While there have been several high-profile scandals around technologies like automated facial recognition,4 predictive policing5 and international mobile subscriber identity (IMSI) catchers6 that have led to their discontinuation, smart surveillance technologies very much like ones used in many smart city projects are still being used by these RTOCs to provide this potentially highly problematic information directly to routine police operations without any public debate or oversight.
Under the guise of “Open-Source Intelligence,” these centres treat all publicly available information on the internet as fair game for surveillance recording. This includes social media surveillance tools that extend police gaze deep into areas of people’s everyday lives in ways that users are mostly not aware of and have not consented to.
While police have long used CCTV in a few key areas, RTOCs enable the deployment of many more cameras by linking existing private (like campuses and malls) and public (like parks, transit, or roads) networks. Research shows that CCTV can exacerbate discriminatory stereotyping,7 and RTOCs enable real-time CCTV access to spaces that previously were not policed in this way.
Another key function is to incorporate ever more databases, from private and public sources, such as the information of people who have merely come into contact with the police (e.g. through programs like “carding”8) to external databases that collect information like demographics, vehicles and licenses and anything else Canadian privacy laws don’t expressly prohibit. Such data mining has well-established potential harms and may even prove to be in breach of Charter rights.9
Smart city technologies frequently utilize the very same kinds of open-source data collection, CCTV coverage, or data mining technologies, yet as the rejection of Sidewalk Lab’s project and the winners of the Smart Cities Challenge show, there is a strong public and governmental push for a public, democratic debate around transparency, harm avoidance and resident participation.
Given the striking similarities and the deep enmeshment of smart city technologies and the RTOCs, there is no reason the same debate should not take place for police utilization of those practices and technologies. As for smart city projects, the Open Smart City principles can provide a strong framework for how to move forward with smart technologies in a way that benefits residents and the communities in which they live.
Open Smart City Principles
Open Smart City principles10 recognize that the benefits and harms of smart technology, in policing as well as in smart cities, are frequently unevenly distributed across communities and as such emphasize values of equity, public oversight and control. These values are distilled in five principles:
The ethical, accountable and transparent governance of the use of technologies;
democratic, participatory and collaborative approach across community, private, public and civil society sectors;
the consideration of technologies that are fit for purpose and that work against harm and bias;
an open data management structure in which privacy is guaranteed and custody and control over data generated by smart technologies is held and exercised in the public interest;
a recognition that data and technology are not automatic solutions to social problems. Rather than quick techno-fixes, these issues often need innovative and strategic social, economic and political solutions.
These principles are the basic requirements for a democratic technological society and local governments are applying them to Open Smart City projects across Canada. Unfortunately, the kinds of opaque acquisition and usage of smart surveillance technologies by police forces as described at the top of this article frequently stand in direct contradiction to them. Indeed, some departments actively worked to prevent legitimate research of their activities!
This conspicuous desire for exemption from democratic transparency and governance says a lot about how these technologies and their usage are understood by police forces. Yet this kind of counterterrorism-inspired approach is counterproductive and many nascent projects around the world have demonstrated alternatives. For example, the Police Data Initiative—which involves more than 120 agencies across the United States—is taking steps in the right direction for open data governance values; in Oakland, California, and now in many other US municipalities, residents have gained considerable power over police technology procurement and budgeting;11 and the Vancouver Police Department has been taking steps to involve many different public and civic groups in the development of its GeoDash predictive policing technology in order to minimize potential harms.
Clearly, these are issues with significant potentially negative consequences for equity and democracy and their resolution should not be left up to the combination of corporations and the police themselves to determine. The Open Smart City principles provide a clear framework of values for public and democratic discussion of their application. A public debate may determine that in certain instances, some of these principles apply differently to policing as opposed to, say, smart urban traffic management, or it may determine that an issue does not warrant a high-tech solution or even a policing solution at all. Whatever the outcome, it is essential that a democratic debate based on common, publicly understood values is had.