Case Study 12.1: The Smart City That Watched Everything

Background

In 2017, Sidewalk Labs — a subsidiary of Alphabet, Google's parent company — announced an ambitious plan to build a "smart city" neighborhood on Toronto's waterfront. The project, called Quayside, would be a 12-acre development on the city's eastern waterfront that would serve as a living laboratory for urban technology. Sensor-embedded streets would monitor traffic flow. AI systems would optimize energy use, garbage collection, and public transit. Heated sidewalks would melt snow automatically. Adaptive traffic lights would respond to pedestrian and vehicle patterns in real time.

The pitch was compelling: a neighborhood designed from the ground up with data and AI at its core, promising to solve some of the most persistent problems of urban life — congestion, energy waste, housing affordability, and environmental sustainability.

But almost immediately, the project ran into a question it could not satisfactorily answer: Who owns the data?

The Promise

Sidewalk Labs envisioned a neighborhood where thousands of sensors would continuously collect data about nearly every aspect of urban life:

  • Environmental sensors would monitor air quality, noise levels, temperature, and weather conditions at a hyperlocal level.
  • Traffic sensors would track pedestrian movements, vehicle flows, and cycling patterns to optimize transportation infrastructure in real time.
  • Building sensors would monitor energy usage, occupancy levels, and structural conditions.
  • Infrastructure sensors would track waste levels in garbage bins, water usage, and the condition of roads and sidewalks.

This data would feed into AI systems that would make the neighborhood more efficient, more sustainable, and more responsive to residents' needs. Traffic signals would adapt to actual pedestrian patterns. Heating systems would adjust based on building occupancy. Snow-melting systems would activate only when and where needed.

In Sidewalk Labs' vision, this was not surveillance — it was optimization. The data was about systems, not people.

The Concerns

Critics were not convinced.

Privacy advocates pointed out that many of the proposed sensors could identify individuals even without collecting names. A network of cameras tracking pedestrian "flows" does not need to know your name to track you — your movement pattern, your height and build, your clothing, your walking speed create a unique signature that can be followed across the neighborhood. Combined with other data sources (phone signals, purchase records, transit passes), "anonymous" movement data could be easily re-identified.

The Canadian Civil Liberties Association (CCLA) filed a lawsuit arguing that the project would create an unprecedented level of surveillance in a public space without meaningful consent from the people who would live and work there. Their central argument: you cannot consent to surveillance simply by walking through a neighborhood. Consent requires a genuine choice, and there is no reasonable alternative to using public streets and sidewalks.

Former Sidewalk Labs advisor Ann Cavoukian, Ontario's former privacy commissioner and the creator of the "Privacy by Design" framework, resigned from the project in 2018. Her reason: Sidewalk Labs could not guarantee that all data collected in the neighborhood would be de-identified at the point of collection. The company had agreed to de-identify its own data, but could not make the same commitment for third-party companies that might operate within the development.

Waterfront Toronto, the public agency overseeing the project, commissioned an independent review that raised fundamental questions about data governance. Who would own the data collected in public spaces? What rules would govern its use? Who would have access? How long would it be retained? The answers were unclear.

The Data Trust Proposal

To address these concerns, Sidewalk Labs proposed creating an independent Civic Data Trust — a new entity that would govern data collected in the Quayside neighborhood. The Trust would:

  • Set rules about what data could be collected and by whom
  • Require privacy impact assessments for all data projects
  • Ensure de-identification of personal data
  • Give residents a voice in data governance decisions
  • Operate independently from Sidewalk Labs

The idea was novel, but skeptics raised questions. How would the Trust be funded? Who would appoint its members? What enforcement power would it have? Could a Trust created by a private company truly be independent of that company? And did the concept of a "data trust" have any legal precedent or enforcement mechanism?

The Outcome

In May 2020, Sidewalk Labs announced it was abandoning the Quayside project, citing economic uncertainty related to the COVID-19 pandemic. Many observers believed the privacy controversy was at least as significant a factor. The project had become a lightning rod for debates about smart city surveillance, and public opposition had made it politically untenable.

The Toronto waterfront site was eventually developed without Sidewalk Labs' involvement, using more conventional urban design approaches.

The Broader Pattern

The Quayside story is not unique. Cities around the world are deploying smart city technologies — sensor networks, AI-optimized traffic systems, predictive infrastructure maintenance — often with limited public debate about the surveillance implications.

In some cities, the deployment has been less visible and less contested:

  • Singapore has deployed a comprehensive sensor network across the city-state as part of its Smart Nation initiative, monitoring everything from crowd density to smoking in prohibited areas.
  • Hangzhou, China uses an AI system called "City Brain" that processes data from traffic cameras, GPS data, and social media to manage urban services. The system reportedly reduced traffic congestion by 15% in its first year.
  • Kansas City, Missouri installed smart streetlights, environmental sensors, and free public Wi-Fi along a major corridor — and initially did so without a public data policy. Community pushback led the city to develop a data governance framework after the fact.

In each case, the central tension is the same: smart city technologies offer genuine benefits (reduced congestion, lower energy use, faster emergency response), but they require data collection that, depending on its scope and governance, can constitute mass surveillance of public space.

Discussion Questions

  1. The consent problem: Sidewalk Labs proposed that people could consent to data collection by choosing to enter the neighborhood. Is this meaningful consent? How is it different from, say, consenting to security cameras in a private store? What about residents who live there and have no practical alternative to using public spaces?

  2. De-identification and its limits: The project proposed de-identifying data to protect privacy. Based on what you learned in this chapter about re-identification and inference, how effective is de-identification as a privacy protection? What are its limits?

  3. The data trust model: Evaluate the proposed Civic Data Trust as a governance mechanism. What are its strengths and weaknesses? Could a similar model work for other AI surveillance contexts, such as police body cameras or hospital diagnostic systems?

  4. Benefits vs. surveillance: The smart city promised real benefits — reduced energy use, less traffic congestion, better public services. At what point do the benefits of data collection justify the surveillance it requires? Who should make that determination?

  5. Lessons for other cities: Based on the Quayside experience, what principles should guide cities that want to deploy smart city technology while protecting residents' privacy? Draft three specific policy recommendations.

Connection to Chapter Themes

This case study illustrates several key themes from Chapter 12:

  • Privacy as power: The core debate was about who would have power over the data generated in public space — a private corporation, an independent trust, or the public itself.
  • The limits of consent: Sidewalk Labs could not solve the fundamental problem that you cannot meaningfully consent to surveillance simply by existing in a public space.
  • Surveillance as infrastructure: When surveillance is built into the physical infrastructure of a neighborhood, opting out becomes effectively impossible.
  • The gap between promise and governance: The technology to collect and analyze urban data exists. The governance frameworks to manage it responsibly are still being invented.