It’s the the final part of my voyage through Part 2 of Chapter 1 of Volume 1. Or, as I like to call it, the Middle of the End of the Beginning of the MIDP.
Previous Master Innovation and Development Plan liveblog entries and relevant documents available here
Part 2.6: Digital Innovation (pp. 230-243; not listed in the Table of Contents)
In this section, Sidewalk Labs begins to set out its data-governance plan, which for the moment consists mainly of general intentions and proposals.
This section is all about Sidewalk Labs setting standards that would give it and Google significant leverage in other markets (standards are power), and defining data in such a way (i.e., a very slippery definition of something called “urban data”) that gives it and other companies access to data generated in Quayside, including data for advertising purposes.
Which Sidewalk Labs has fought every step of the way
Toronto and Ontario have taken some important initial strides to advance the conversation around data governance principles. (p. 232)
Toronto Star headline, August 14, 2018: “Sidewalk Labs unveils plans for timber towers, raincoats for buildings in Quayside, but Torontonians must wait for data details”
Key components of its data-governance plan
Distinction between urban and non-urban data: your regular reminder that urban data is a term that does not exist in Canadian law, leaving Sidewalk Labs with a lot of leeway to define it in its own interest.
An “open digital infrastructure” that will use “urban data to improve quality of life.” This would involve a “standardized mount system” that would “eliminate vendor lock-in” and would allow Sidewalk Labs a shot a setting a global standard. (p. 233)
“Open and secure” data standards. A set of publish standards around open-data architecture, access and sources”
Goal: to allow third parties to build on Sidewalk Labs’ system (i.e., creating a platform) (p. 233)
Question: is Sidewalk Labs’ standardized mount, “Koala,” a proprietary design?
New bureaucracy: An Urban Data Trust
anchored by a Responsible Data Use (RDU) Assessment — an in-depth review that is triggered by any proposal to collect or use urban data — and guided by a set of RDU Guidelines that incorporates globally recognized Privacy by Design principles. (p. 233)
- The Urban Data Trust would be “independent,” “not controlled by either Sidewalk Labs or Waterfront Toronto”
- It would have a five-person board and a Chief Data Officer
- It would approve “all collection or use of urban data in Quayside”
- It would only cover “urban data” (other types of data are not covered by this section) (p. 240)
More details apparently are in Volume 2.
It would be responsible for establishing “a set of Responsible Data Use (RDU) Guidelines that would apply to all entities seeking to collect or use urban data in the IDEA District, incorporating globally recognized Privacy by Design principles.” (p. 240)
Sidewalk Labs would play a key role in setting these policies, which, again, would be based on a category “urban data” that it has defined.
According to Sidewalk Labs, this policy should be governed by the following principles:
- Beneficial purpose for urban data’s use;
- transparent and clear notification to individuals about how this data would be collected.
- Minimize data collection with the least-invasive technology
- Publicly accessible data by default (property de-identified or non-personal data). No matter that it’s becoming increasingly clear that truly deanonymizing data is an impossible task.
- No selling or advertising without explicit consent. Of course, urban data by definition is collected in public spaces where it is pretty much impossible to get valid individual consent. So, consent would have to happen in another way; i.e., via mass consent.
- Here’s the problem: Smart cities require ubiquitous surveillance and data collection. Getting individual consent for data collected from an individual’s movement through a public space is pretty much impossible. So consent would have to be done in a pre-determined manner (e.g., by merely entering the space, or by posting a sign, or). In other words, the Urban Data Trust is the only organization that can define what “mass individual consent” actually entails. It’s not going to shut down the smart city. Rather we’ll find that getting legal (but ethically fraudulent) “individual consent” can be achieved simply through a bit of bureaucratic manoeuvring, of defining the problem away. This is a good example of how Sidewalk Labs’ commitment to actual individual privacy is much less than it seems.
- Responsible AI principles (all these on p. 241).
It would also
implement and manage a process for approving the responsible collection and use of urban data anchored by a publicly auditable Responsible Data Use (RDU) Assessment — an in-depth review that is triggered by any proposal to collect or use urban data. (p. 240; proposed rules listed on page 241)
Urban data produced by Sidewalk Labs’ platform would be “publicly accessible, enabling companies, community members, and other third parties to use it as a foundation to build new tools.” (p. 233)
- How would this use be qualified?
- To what extent would this involve the possibility of privatizing public services?
- Since this data would be produced from (Canadian/Toronto) residents and Toronto geography, will Canadian governments, NGOs and/or companies have priority access to this data? Otherwise, one could easily anticipate the monopolization of this level of services by data giants like Google.
- Who will maintain the platform? (Sidewalk Labs, I think.)
- For how long?
- What happens after the first contract finishes?
- What will the public authorities do to ensure that the city isn’t held hostage to Sidewalk Labs as the keeper of the platform?
Its promises for “best-in-class resiliency and security” are just that – promises – unless and until we see exactly what they are doing, beyond their current list of things the company (which isn’t a city) currently uses (p. 239).
What’s not here
What about “non-urban data”? Maybe it’ll be discussed later?
Public Engagement (pp. 244-249)
Nothing to see here: This part is repeated word-for-word from the Overview.
We have a Very Special Episode for Post 20: a quick look at the economic analysis underlying Sidewalk Labs’ projections and we all learn an important life lesson. See you there.