If you care about corporations monetizing data generated by people’s desire to communicate, you may have read worrying stories featuring some of the Google octopus of companies, including Sidewalk Labs, whose ‘smart city’ proposals for Toronto’s waterfront have drawn much critical attention in this blog and elsewhere.
The latest Google strategy is to invent a category previously unknown to data science: “urban data”. It appears to respond to citizen concerns about data mining, while in fact laying the groundwork to allow Google companies to continue to garner huge profits and evade government regulation.
On April 16, the Canadian Civil Liberties Association announced a lawsuit against Waterfront Toronto (WT). In non-technical terms, the lawsuit argues that Waterfront Toronto never had the authority to issue a highly indefinite call for proposals inviting private companies to plan a new district full of novel data collection mechanisms. In response to the lawsuit, Google affiliate Sidewalk Labs (the company awarded the right to develop such a ‘smart city’ proposal) issued a public statement saying that it cares greatly about privacy in the collection of “urban data”.
On April 19, Sidewalk Labs (SWL) released a set of statements about transparency in “urban tech”. Their statements were developed by workshopping a set of 94 slides. The main goal of the workshops at which the slides were shown was a peculiarly narrow one: to come up with standardized icons warning citizens that data are being collected in certain places. The drawn-out exercise was described as setting standards for “transparency”.
Transparency is of course the very virtue that neither SWL nor WT have shown since the ‘smart city’ plan was first announced, in October 2017. But in a webpage of uncertain status called “Sidewalk Labs Talk”, Sidewalk’s Jacqueline Lu writes: “How can we bring transparency into urban tech? These icons are a first step.” (For more on the icons and their Toronto-based production process, see Shannon Mattern’s informative article.)
In the slide deck SWL gives many examples of the sensors that Sidewalk would bring to public notice via the icons. But as computer scientist and Waterfront Digital Strategy Advisory Panel member Andrew Clement noted in an April 23 talk at the University of Toronto, the company’s pictures of future sunny public spaces by Lake Ontario (and the slide deck just mentioned) rigorously omit to mention the data we create for the private sector when we walk around with ‘smart’ phones. The sensors in your pockets, as Clement called them, are wholly invisible in all the Sidewalk pretty pictures.
This sleight of hand is achieved by their self-generated definition of “urban data”. “Sidewalk Labs has defined Urban Data is [sic] data collected in a physical space in the city, which includes public spaces such as streets… and parks… [as well as] private spaces accessible to the public such as …retail stores.”
The same slide makes the astounding statement that in the Google universe only homeowners have privacy rights: “Urban Data”, Sidewalk continues (capitalizing “urban” and “data” as if to trademark the words), includes not only data collected from sensors in public spaces but also data collected in “private spaces not controlled by those who occupy them (e.g. apartment tenants).”
So, in the future ‘smart’ Toronto neighbourhood, landlords would be free to place sensors and surveillance cameras inside their tenants’ apartments, but “transparency” would be achieved by the tenants seeing, on their kitchen walls, the ‘you are being recorded’ icons that Sidewalk has been workshopping. (In fact, tenants and everyone else in Quayside would get some protection from the current federal privacy law, PIPEDA, and the municipal counterpart, MFIPPA; but in its “urban data” materials, Sidewalk Labs does not acknowledge existing laws).
But even homeowners would be exposed to having their data mined as they use the attractive public and semi-public spaces that Sidewalk wants to build. Our phones reveal the owner’s location every few seconds, sometimes even if the GPS is turned off. And as we know from many recent investigations, companies are getting better at collecting and monetizing data whose collection does not breach federal or other privacy law because the data are notionally not personally identifiable.
Excluding cell phone location (and all manner of data generated from using laptops in cafes or other public spaces) from the definition of “urban data” gives the Google companies free, unregulated access to troves of useful information about how certain types of people move across the city. National or local digital policies and data privacy law might restrict such data gathering, after the fact; but existing laws are geared mainly to governing the collection of personally identifiable information by purpose-built sensors and cameras (such as surveillance footage that shows your face).
The expensively facilitated “transparency in urban tech” consultation SWL carried out in March and April focused strictly on sensors that are nothing but sensors, such as traffic cameras, licence plate readers, bank ATM cameras, and Presto card readers. By contrast, laptops and phones (as Prof. Andrew Clement showed to great effect in his presentation) are strangely absent both from the slide deck and in Sidewalk’s ‘smart city’ by Lake Ontario drawings. To check Clement’s point, in a data-policy game of “Where’s Waldo” I scrutinized countless images on the Sidewalk website, looking for someone making a phone call. Nothing. And neither laptops nor phones are mentioned in the “transparency in urban tech” slides. Look at the icons, SWL is telling us, don’t think about your phone and what it’s doing even as you offer opinions about whether this icon should be yellow or red.
Another major lacuna in the SWL images, and also an absence in the slide deck mentioned, is street advertising. SWL's subsidiary, Intersection, is a major player in the digital out-of-the-home (aka DOOH) advertising market and has an impressive and disturbing range of offerings that are increasingly reliant on exploiting personal data from mobile devices and other sources.
A final point is that Sidewalk has been promising to benevolently place data collected at Quayside into a ‘data trust’. But is the data trust intended only for “Urban Data”? The data trust is not mentioned in any of 94 “transparency in urban data” slides, so this cannot be confirmed, but there must be some coordination between the data trust idea and the “Urban Data” exercise in defining what data collection processes are to be governed at all. If the trust is indeed intended only for the mainly publicly collected “urban” data Google wants us to notice, Torontonians will waste their time debating who should govern the data trust while Google continues to mine everyone’s cell phones and laptops for data that is not “urban”, data Google can take all the way to the bank.