Twitter
Facebook
YouTube
LinkedIn
RSS

Panel 218 - Digital Futures: The Impact of Digital Threats to Democracy

Conference Day: 
Day 3 - November 9th 2018
Takeaways and recommendations: 

Digital futures: the impact of digital threats to democracy

Organized by: Social Sciences and Humanities Research Council, Eloisa Martinez

Speakers: Dr. Elizabeth F. Judge, Professor of Law at the Faculty of Law, University of Ottawa; Barrie Kirk, Executive Director, Canadian Automated Vehicles Centre of Excellence (CAVCOE); President, Canadian Automated Vehicles Institute (CAVI); Norman Mendoza, Manager of Business and Technology Innovation, City of Edmonton; Renee Sieber, Associate Professor, School of Environment & Department of Geography, McGill University

Moderator: Ursula Gobel, Associate Vice-President, Future Challenges, Social Sciences and Humanities Research Council

Takeaways and recommendations

The opportunities and risks of algorithmic governance

  • The world is moving increasingly towards algorithmic governance, which entails automated decision-making based on massive data collection. These automated processes could be used to make decisions about a wide range of human activities, including credit, insurance, adoption, employment, jury selection and sentencing, and parole decisions.

  • The potential advantages of algorithmic governance would be a more individualized system, like personalized medicine, that could mitigate the risks of arbitrary and subjective human decisions. However, the potential downsides to algorithmic governance would be a Kafkaesque-like situation, where faceless decision-making would remove the best of human decision making (e.g., mercy, flexibility, discretion) and could be based on inaccurate data or faulty premises.

  • Algorithmic governance is a potential threat to democracy (i.e., it can replicate the worse of human decision-making, reflect biases, violate privacy, limit freedom and take away our ability to influence government).

  • Creating transparency around these algorithms and how they make decisions is hampered by trade secret protection for algorithms under intellectual property laws that lead to informational asymmetry. “Algorithms know more about people then people know about the algorithms.”

The opportunities and risks of autonomous vehicles (AVs)

  • AVs will have sophisticated roaming sensor platforms that can monitor people in the vehicle and on the street – data that can be uploaded in real time, analyzed and ultimately monetized.

  • This large scale surveillance of the population can affect free speech (i.e., police using non-passenger roaming AVs with facial recognition capabilities).

  • AVs and driverless taxis will improve mobility for the elderly and disabled, enabling them to better engage in society and civic activities like voting.

  • The biggest threat is the federal government. There are several reasons why governments cannot manage disruptive technologies:

    • Pace of technological change

    • Scope of technological change

    • Risk culture vs risk adverse

    • Few borders

    • Platform-based / near zero marginal costs

    • Trial and error vs impact of uncertainty on government policy

    • Disruption by interactive social-media platforms with enormous scale

  • Canada could benefit from using quasi-autonomous, non-governmental organizations (QUANGOs) to manage disruptive technologies. QUANGOs in the UK are funded and overseen by government, but much of the decision making is done by industry.

  • Several federal departments are involved in policies related to AVs, but there is too much duplication and not enough harmonization.

The role of governments in open data

  • The Canadian Geospatial and Open Data Research Partnership (Geothink) looks at the implications of increasing two-way exchanges of locational information between citizens and governments and the way in which technology shapes, and is shaped by, this exchange.

  • One of the biggest issues for policymakers is how to keep data open and accessible in the future.

  • Canada is making good strides in open data. It co-chairs the Open Government Partnership which aims to create and sustain open data.

  • Open data is supposed to make government more transparent and accountable, but that doesn’t always happen. It depends on:

    • Availability: It’s not enough to just publish data. You have to deliver value. “Info-mediaries” (perhaps QUANGOs) could create value, through digital literacy among the public. You also cannot connect open data to democracy without a legal regime that holds governments accountable for what is found in the data. Transparency does not automatically equal accountability.

    • Economic development: Open data is increasingly being operationalized for economic development. Fairness and equity must be paramount to ensure all citizens benefit.

    • Substitution: Increasingly, government is accepting data from the public (i.e., crowdsourcing) and even substituting public data for data it used to collect. But that data comes with its own biases and assumptions.

  • Rather than looking at the impact of digitalization on democracy, look at the impact of democracy on digitalization.

  • Many jurisdictions are guided by the principal of “open by default” which emphasizes quantity over quality – “just get the data out there”. Many discussions are returning to the idea of “publishing with purpose” – release the open data that can have the most significant impact.

  • Open data has become tool to regain trust in many jurisdictions that have suffered from government corruption around world. For countries like Canada, open data can demonstrate that trust in government is well placed.

  • Open data doesn’t guarantee government accountability but it can provide the evidence that makes the need for accountability harder to ignore.

  • Municipalities are leveraging open data to deliver new services or improve services. For example:

    • The City of Edmonton discontinued its transit app because companies were able to provide riders with a much better service using real-time transit data.

    • Another popular Edmonton dataset is a list of trees maintained by the city that bear edible fruits. Citizens are allowed to pick the fruit at the end of the season, saving the city the time and cost of cleaning up rotting fruit.

  • Open data is not a substitute for full citizen engagement, but it can help governments improve public services.

Related reading:

Documents: