Technologists for Democracy logo

Author: Khasir

  • Annex Gleaner, Aug 2025

    Volume 31, Number 3 (full text follows)

    Self-driving delivery vehicles spark privacy concerns

    Province approves Magna pilot for several wards including University-Rosedale

    By Abigail Harris

    Magna International, a global automotive supplier and mobility tech company, launched their Automated Vehicle Pilot Program in May 2025.

    The program is testing electric, self-driving delivery vehicles (SDVs) as part of a food delivery service, and the pilot area includes the Annex.

    The Ministry of Transportation Ontario (MTO) approved the pilot in April, allowing the vehicles to operate in Ward 9 and in portions of Wards 4, 5, 11, and 12.

    All vehicles will receive their own Ontario licence plates and will be subject to the same road rules as conventionally operated vehicles.

    According to Magna’s website, sensors on the vehicle will collect data from the surrounding area which will “incidentally include personal information such as images of people and vehicle license plates.”

    In June, Khasir Hean and Ilya Kreynin, members of the group Technologists for Democracy, wrote a letter to Magna along with four other members detailing their privacy concerns with the pilot.

    Kreynin and Hean joined other Technologists for Democracy members at city hall in June to discuss their concerns with the city’s infrastructure and environment committee, and city councillors passed a motion on June 26 to investigate data privacy issues surrounding Magna.

    Kreynin said that they became aware of the issue when fellow member JJ Fueser told them about the new self-driving vehicle pilot.

    “When JJ (Fueser) first brought this to our attention, and we gave a brief read of the open privacy policy that Magna has, it was egregiously overstepping,” Kreynin said. “We realized that there was really low awareness of this and that there wasn’t anybody else who was speaking out about this directly.”

    Hean said Magna is recording faces because they believe them to be useful in determining the gait of passersby in relation to their vehicles.

    “Magna fully intends to record the faces of individuals that it passes by on the streets because it believes that faces are useful to help predict where people are walking,” Hean said. “Other self-driving vehicle companies don’t record people’s faces. They blur them.”

    Magna states that they will retain this data for “as long as nec-essary,” which Kreynin says sets a dangerous precedent.

    “That’s entirely vague. They can keep it however long they want,” Kreynin said. “This is a new level of privacy breach in terms of this type of storing and anonymization. This company is running their cameras and robots into our neighborhood where people did not choose and consent to this happening.”

    The privacy policy also asserts that data will be shared with third parties such as insurance providers, law enforcement, and other officials when required by law.

    Hean says this raises many questions about Magna’s process of sharing this information and its ethicality.

    “Where’s the accountability in that?” Hean said. “How will we be informed if the police decide to use Magna as a live tracking device as well as a food delivery service? How do we and the public know this? There’s no mechanism to inform the public.”

    Kreynin says Magna responded to their letter in July, not by addressing their concerns, but by inviting them to tour their facility.

    [Image of a man picking up food from the Magna self-driving vehicle, with caption: Magna’s pilot program supports small-package deliveries like food, groceries, and retail items.]

    “While they invited us to the facility, they have not responded to any of the clear and direct concerns laid out in the letter in writing,” Kreynin said. “We have not received any kind of satisfactory or clear response from Magna at this time.”

    According to Kreynin, an important consideration in this emerging new technology is whether it is helping or hurting communities.

    “The point of the vehicles is to replace workers,” Kreynin said. “This industry is only set to expand in a way that is likely to hurt local businesses and communities. The question is, what is the effect that this has on our society? This pilot is not just a pilot [of] technology; it’s a pilot of integration into our society.”

    Magna International did not respond to the Gleaner’s request for an interview.

  • Letter to Competition Bureau on Algorithmic Pricing

    In response to the Competition Bureau Canada’s call for feedback on algorithmic pricing, we submitted the following letter.

    See original PDF here.

    August 3, 2025

    Re: Algorithmic Pricing and Competition

    Algorithmic Pricing Needs Greater Transparency and Guardrails

    Dear members of the Competition Bureau,

    We are Technologists for Democracy, a grassroots advocacy organization based in Toronto. Our position largely concerns the effects of algorithmic pricing from the consumer perspective.

    Most consumers – 68% – feel that dynamic pricing unfairly takes advantage of them.1 This is not necessarily because of an inherent unfairness of dynamic pricing, but because most companies operate dynamic pricing through a purely profit-oriented lens focused on the short term. There is little thought given to transparency, company trust, or consumer well-being.

    Consumers are particularly vulnerable to downstream effects of algorithmic pricing, especially in markets of essential goods, markets with high barriers to entry and effective monopolies. Without reasonable alternatives, consumers become captive to price increases. Low-income consumers are particularly affected, due to any given purchase constituting a larger portion of their income. Housing is one such essential market – we welcome the Competition Bureau’s current probe into the use of algorithmic pricing for setting rental prices, and would recommend prohibiting the use of algorithmic pricing for housing (recommendation #6 below).

    In addition, we believe algorithmic pricing can lead to market inefficiencies. When companies offer personalized dynamic pricing, with different prices for each individual consumer, it becomes difficult for consumers to compare and recommend prices between competitors, and difficult for competitors to efficiently set prices according to the market. We see personalized dynamic pricing as a dangerous opportunity for larger entities to unfairly exploit their market dominance by reducing the information available to their competitors.

    To limit the harmful downstream outcomes of algorithmic pricing, our recommendations focus on improving transparency and guardrails surrounding algorithmic pricing. In the same way that nutrition facts labels and ingredient lists allow consumers to make informed decisions before purchasing food, labels on algorithmic pricing would allow consumers to make informed decisions before purchasing digital and digitally-enhanced products and services. 

    To improve transparency, we support regulation and enforcement which would require that companies clearly disclose:

    1. Whether or not algorithmic pricing is in effect for a given product or service.
    2. If algorithmic pricing is in effect, whether the pricing model was developed in-house or is outsourced to a third party.
    3. If algorithmic pricing is in effect, whether or not AI/machine learning is used for algorithmic pricing, as opposed to a rule-based model.
    4. If algorithmic pricing is in effect, what data is inputted into the pricing model. For example:
      1. Consumer data such as location, credit score or demographic profile,
      2. Inferred data such as consumer emotional state,
      3. Internal data such as sales counts,
      4. External data such as competitor prices or current weather,
      5. Etc.

    To ensure enforcement of policies, we support: 

    1. Establishing team(s) and process(es) to handle complaints and appeals for policies related to recommendations in this letter.

    In addition to educational regulation, we also support investigations as to the feasibility of:

    1. Prohibiting personalized, dynamic algorithmic pricing from being used in certain market sectors such as those of essential goods and services (including food, housing, and medication).
    2. Regulation of prices such as through a Maximum Retail Price policy, effective in countries such as India.

    We believe that these suggestions will:

    1. Improve consumer trust of algorithmic pricing.
    2. Inform consumers as to what personal data is involved in making a purchase.
    3. Lower the competitive barrier to entering markets already populated with algorithmic pricing models, while still allowing companies to maintain secrecy of the inner workings of proprietary pricing models.
    4. Reduce possible harms on consumers and competitors alike by preventing predatory pricing.
    5. Ensure equal opportunity of access to essential goods and services for consumers.

    We urge the Competition Bureau to increase transparency and implement guardrails on the use of algorithmic pricing.

    Sincerely,

    Khasir Hean
    [email removed for privacy]

    Henry Wilkinson
    [email removed for privacy]

    Jenny Zhang
    [email removed for privacy]

    Cole Anthony Capilongo 
    [email removed for privacy]

    Technologists for Democracy
    techfordemocracy.ca

    1. Gartner Marketing Survey Finds 68% of Consumers Report They Feel Taken Advantage of When Brands Use Dynamic Pricing. December 16, 2024. https://www.gartner.com/en/newsroom/press-releases/2024-12-16-gartner-marketing-survey-finds-68-percent-of-consumers-report-they-feel-taken-advantage-of-when-brands-use-dynamic-pricing ↩︎

  • Stop Bill C-2 from Surveilling Us

    May 2025 — October 2025

    Update 26/04/29: As of Oct 8, 2025, Bill C-2 was put in limbo and replaced with Bill C-12, which removed the privacy invasions but kept many of the problematic anti-immigration policies. Bill C-12 passed into law on Mar 26, 2026.

    Update 26/05/04: Bill C-22 was introduced in March 2026 as a reworked version of the surveillance aspects of Bill C-2, and continues to threaten privacy rights and civil liberties.

    Tell your MP to say NO to Bill C-22!

    Our archived post is below.

    Bill C-2 was first introduced in May 2025 as an omnibus bill. It has faced criticism by multiple organizations:

    Many organizations have also criticized other aspects of the bill:

    There are a number of concerns that we as technologists have with regards to Bill C-2:

    1. Loss of privacy from requiring electronic service providers to provide basic subscriber information to CSIS and law enforcement agencies without a warrant.
      • Creation of a new term, electronic service provider, which includes social media platforms, email providers, messaging services, gaming platforms, internet service providers, and more…
    2. Reduction of oversight over CSIS and law enforcement agencies.

    Take action now to stop Bill C-2!

    Contact Your MP!

  • Tell Magna to Stop Recording our Faces

    June 2025 — October 2025

    Update 10/02/25: Magna is no longer operating its pilot in Toronto! Thank you to everyone for your support in making this happen. Innovation does not need to come at the cost of privacy!

    Our original post is below.

    On May 26, 2025, Magna International Inc. began a self-driving vehicle pilot for food delivery within the City of Toronto. Magna intends to blur license plates but record people’s faces within the vicinity of the vehicle because “Magna believes that faces are very useful to determine gaze, pose and intent of motion” according to a background report by the City.

    Other companies like Real Life Robotics, which is running a self-driving vehicle pilot in Markham, aren’t recording people’s faces: “the AI software [Real Life Robotics] use can easily scrub identifying features, like faces and licence plates.”

    How you can help!

    On June 27, we sent a letter (read here) to Magna with a list of specific privacy concerns based on their privacy policy.

    If you want to get involved with strategy and planning, sign up on our volunteer page!

    Get Involved!

    More info

    A view of Magna's delivery vehicle from the back. The vehicle is smaller than a car and has three wheels.
    Magna’s last-mile delivery device (LMDD) operating in Michigan.

    On June 11, 2025, the pilot was discussed at the City of Toronto’s Infrastructure and Environment committee. The pilot is under the jurisdiction of the Province of Ontario, but Toronto is allowed to submit comments and feedback to the Province.

    A map of the pilot area in Toronto.
    The pilot area in Toronto covers all of ward 9 and parts of wards 4, 5, 11 and 12.

    JJ, Ilya, Khasir and Akhil spoke up about the privacy concerns surrounding the pilot. Thanks to their voices, city councillors passed a motion on June 26 to investigate data privacy issues through the following questions:

    1. What type of data is being collected?
    2. Who has access to this data?
    3. Where is this data being stored?
    4. Can or will this data cross the border into the United States?
    5. Can Magna entities within the United States view the data?
    6. Would Magna entities be required to provide data to US or Canadian law enforcement if requested?
    7. Will facial recognition data be shared with law enforcement?
    8. Can future agreements change the nature of how the data is collected, stored, transmitted, and used?
    9. Are there any mechanisms for individuals, including those who cannot provide consent, such as children, to opt-out or request that their records be deleted?
    10. Are there ways to ensure that data can be destroyed if no consent was given?
    11. Can the terms and conditions of data management be changed, and if so, by whom?
    12. Does the City’s Digital Infrastructure Strategic Framework govern the collection and use of this data and if not, what can the City do when third parties want to harvest data in the public realm?