2020/01/22
==========

* Sidewalk Labs Privacy Panel
* Is Privacy Possible in the proposed Sidewalk Toronto or any "Smart City"?

- Andrew Clement, Professor Emeritus, Faculty of Information, University of Toronto and Co-founder, Identity Privacy and Security Institute (IPSI)
- https://www.ixmaps.ca/

- Natasha Tusikov, Assistant Professor, Social Sciences, York University, Author of Chokepoints: Global Private Regulation on the Internet

- Brenda McPhail, Director - Privacy, Technology and Surveillance Project for the Canadian Civil Liberties Association

- Brenda intro
  - sidewalk labs characterizing the project as 'never being about data'
  - is privacy actually possible in this or any other 'smart city'?
    - given the data driven analysis and embedded sensors issue
  - to start, looking at the Sidewalk Labs consent model
  - how do the Quayside consent documents lay this out

- Natasha
  - Sidewalk Labs has been pushed into having to consider consent and data
    governance by the concerted efforts of activism
  - Sidewalk Labs documents trying to collapse the notions of public and
    private data because 'it's hard to obtain consent'
  - the plan is to put up signs with symbols and they want to claim
    they constitute consent
  - apparently Sidewalk Labs has dreamed up these symbols on their own
    as well

- Brenda
  - how will the embedded sensors work and what are your concerns

- Andrew
  - the conflation of smart and intelligent by putting digital processing
    inline is a distortion of language
  - the signage (or lack therof) for video surveillance in the city hardly
    is sufficient for consent, if you need to walk down the street you
    need to walk down the street even if it has signs warning about cameras etc
  - sidewalk labs is extending the operation of currently existing sensors
    such as keycards to enter your building - we don't think about the
    logging that already occurs
    - Ring exists, id scanning for the Winnipeg liquor store exists, Nest
      exists, etc
    - the idea (Andrew sists on the WT advisory board) is that the data
      collected will make it more efficient to manage this area of the city
    - but there hasn't been enough discussion on how to protect the data
      in the interest of the individuals
   - the 'data trust' language of a 3rd party protector or holder of the data
     has been replaced with the new language of a 'collaborative data hub'

- Natasha
  - IoT can be both industrial or consumer data collection devices
  - the software is important, who embeds the rules into the software
  - clearly the original company, but the data is also distributed to
    3rd parties - such as the 3rd party cloud providers, app writers
    - where is the data stored, how long will it be kept, who has access?

- Brenda
  - does it get more complicated with the additional involvement of 3rd
    parties, developers, etc?

- Andrew
  - there's a much wider scope of use and abuse but Sidewalk having this
    discussion offers an opportunity to develop our relationships with
    data collection devices out in the open instead of these devices
    infiltrating our lives piecemeal
  - Sidewalk now has the responsibility to set data governance rules in
    conjunction with the three levels of government
  - Waterfont Toronto is planning on approving the MIDP in March when they
    meet

- Brenda
  - in the face of the existing policy deficit, the data handlers will claim
    that the data will be deidentified and aggregated, is this an effective
    privacy protection?

- Natasha
  - skeptical of the claims since there is a growing body of research on how
    to reidentify people through combining multipe sources of information
  - Sidewalk labs claiming no harm can come from deidentified information
    but there can still be personal effects on people from the bias built
    into garbage ML algorithms
- Andrew
  - PIPEDA says if you need to use deidentified data, there are rules
  - however Sidewalk seems to believe deidentification is a get out of jail
    free card, thinking more from an American legal context

- Brenda
  - we've laid a foundation here that suggests privacy is going to be really
    hard in this neighbourhood
  - one feature of the neighbourhood is supposed to be that it will offer
    some affordable housing
  - what does it mean to encourage poor people to live in these neighbourhoods
    when they may not have many other choices in this city

Natasha
  - if you want to look at the future of surveillance, look at the poor
  - little opportunity to opt out if you want the job, if you want the
    home, if you want into the country
  - terrible message to send when affordable housing elsewhere doesn't
    require this
  - sidewalk labs - "if you don't like it you don't have to come here"

Andrew
  - if Quayside is going to do this, who says the whole area under their
    control needs to be blanketed in sensors
  - if this is a serious testbed you need to instrument the place with a
    control area with no sensors for a proper experiment
  - I don't accept the premise then that all of Quayside has to be a
    surveillance zone
  - what if Sidewalk made a promise to its visitors that they could
    visit anonymously, leave data collection as an option for those
    who may live there and want it
    - options like this have not been part of the conversation or the design
      discussion

Brenda
  -

Natasha
  - the Sidewalk/Waterfront partnership has no credibility, the original
    RFP explicitly said the winning partner could write its own rules
    - a tech vendor gets to write the regulatory framework to govern
      its own commercial venture

Andrew
  - if WT wants to regain its role as a reputable public steward since
    they are not a monolitich organization
  - however we should consider Sidewalk and Google to be the same organization
    and seems much of their effort here is public relations and they're
    willing to change their language but there is no evidence they've had a
    change of heart
  - it would have been better if we'd taken barcelona's approach with broad
    public engagement
  - can Sidewalk be saved? well life is full of shabby compromises
  - Sidewalk has no experience in this either, apart from the LinkNYC spy
    kiosks whay are we taking this seriously
  - Sidewalk needs Toronto, this is the only thing they have going it's PR
  - Waterfront Toronto needs Sidewalk for a 'win' as well this is an unholy
    alliance
  - cities are built from messy processes, maybe the best to hope for is that
    we'll learn to govern better as a result

* Audience Q&A

- why can't we have a participatory process?
  - Andrew: if you want to stop this train Toronto people need to effect change
  - Brenda: Support CCLA's lawsuit or try to influence city council, they're
    discussing their own data governance laws tomorrow morning

- qua slippery slope, where are the red lines that might stop the ongoing
  hurdle of Sidewalk Labs
  - Natasha - can WT in cooperation with other regulators actually set rules in
    place for consent and data ownership
    - SL has pivoted to common ownership of the available data but what
      actors have the analytical capability to actually mine this data
    - rich tech companies for the most part
  - Andrew - there are a number of places people might look to intervene or
    contribute to this project
    - Canadian businesses and Council of Canadian Innovators for example, are
      saying it is going to be harmful because all the sector activity will
      be controlled by Google, you could make alliances with Canadian business
    - first nations groups have articulated a set of principles called OCAP,
      around control of cultural artifacts and this maps onto data
    - a good place to start would be to say that those contributing to a
      collective asset should be the ones benfitting from it
   - the MIDP is the glossiest, most seductive peice of print I've seen we
     need to get imaginative and think up better ways to have informed
     relationships with our devices

- Waterfront Toronto shouldn't have rules about digital governance at all
  they are an appointed board to manage property, it's not their job they
  have an appointed board
  - we have elected representatives, look to democratically elected bodies
    that is what they are for
  - democracy to me is more important than privacy we should hold our elected
    off

- about your observations around sidewalk labs signage, how can we even trust
  that Sidewalk will stick to keeping signage up when many Toronto police
  cameras don't have signs any more?
  - Brenda: Google is trying to set standards which if they're succesful
    will be a business model (IP?) for them and will shift the burden to
    those in the area to read understand and consent to surveillance

- the staff report for city council which is up for approval in 2021 has
  five broad princples for digital governance on which law is expected to be
  drafted, for anything that comes up before then (MIDP) the idea is to apply
  these broad principles
  - Andrew: it's a whiff of a good idea but these principles are too thin,
    they're not ready to make decisions that are consequential, they're
    a joke like "well run city" they're not even full sentences they
    derive from Amsterdam, Barcelona, so have promise but it would be a
    disservice to democracy to use these, they're necessary but far
    from sufficient

- what are the international implications of my data, will i see my data
  if I go to another smart city
  - Natasha: keep in mind this is a test bed, it's more the idea that what
    we do here will be echoed internationally
    - smart city activists internationally are also watching this very closely
      since what Toronto ends up here may be replicated internationally

- followup, are there any good examples to look to? to use as a roadmap
  - Natasha: Barcelona is the classic example of a bottom-up design in which
    people asked what existing problems could be solved by tech rather than
    just pulling in a big tech company to tell them what they need
 - Andrew: There is no roadmap,

- there is a level of learning we need to even be able to participate in
  this conversation, what do you look to to keep up with this, public
  intellectuals, podcasts, etc
  - Natasha: Cory doctorow, Motherboard, Jenny Morisov at the Guardian,
    Rob Kitchen, Shoshana Zuboff
  - Andrew: the point is to develop a language that's accessible, I
    would like to hear if people understand something or not so I
    can reframe ideas or make them more accessible, I can sometimes
    overconceptualize thingsbut we need to find modes of talking about
    privacy that is grounded on people's concerns
  - Brenda; Bianca Wylie and BlockSidewalk and the Good Jobs for All
    are examples of community activists that are engaging with the community
    on issues that are comprehensible and speak to local issues

- why is Waterfront Toronto tasked with helping to set rules that may widely
  affect Toronto, Canada, or possibly the rest of the world?
  - Andrew: Waterfront Toronto may not have gotten into this entirely of its
    own volition, they've discovered a cuckoo egg in their nest, and
    probably looked around and said "oh shit" once they realized what
    they'd gotten themselves into
    - maybe a good lesson from this is that if you think you can divorce
      the internet from the real world, this is a mistake. we have to develop
      a way of thnking and planning and discussing this now that sensors
      are being interweaved into the physical world

- how prepared are authorities like the privacy commisionner to deal with
  the coming challeneges of sensor
  - Natasha: the privacy Comissioner is very aware of the risks and
    particularly since cambridge analytica has been pushing to strengthen
    privacy laws
  - Andrew: in some ways our privacy laws and commissions are in some ways
    artifacts of an earlier time when you could identify where the databases
    were but now your information is trafficked and dispersed in untrackable
    ways and the powers hey do have are based on complaints, and that is why
    the laws don't work we shouldn't have to depend on citizens to exercise
    their rights this way