Thank you for the opportunity to be here today. I would like to begin by acknowledging I am speaking from unceded Indigenous lands. The Kanien’kehá:ka Nation is recognized as the custodians of the lands and waters on the place I call home.
I am a professor in Information and Communication Technology Policy in the Department of Communication Studies at Concordia University.
My comments, in a short time, aim to cover my broad, multidisciplinary research into democracy and digital technology based on 20 years of direct experience and historical research dating back to the 1950s. I have written three books broadly related to AI and democracy, two of which are open access, Internet Daemons and my new book SimPolitics: America’s Quest to Solve Politics with Computers published this year by MIT Press.
I have three key recommendations:
1. Create a Minister of Democratic Institutions and AI
I have participated in at least three consultations to address the social impacts of AI technologies. None of these efforts have merited a serious response from policy makers, and too often, meetings are simply used as evidence of consultation.
I cannot be more clear. Canada’s approach to AI governance is undermining its democratic governance. Time and time again, hastily developed industrial strategies with vague promises of digital sovereignty undermine actual democratic sovereignty.
Canadians and Canadian civil society have answers and solutions if they were able to meaningfully participate. In 2023, I was part of the Art-IA strategy, an internationally recognized collaboration on commons-based management solutions for AI in the cultural sector [Figure 1]. That 18-month consultation brought together 203 participants across the arts and culture sector in Quebec, only to present once before committee. We’re still waiting on movement from the government on meaningful policy for AI in the cultural sector.

Figure 1. Art-IA recommendations
In 2025, I was lucky to be part of a broad coalition around the People’s Consultation for AI. Following an open letter with over 200 signatures, the Consultation held a 4 month open call that received 66 submissions. Thanks to my colleague, Dr. Joanna Redden, Meghan Voll, and Leo Litke we understand the range of submissions [Figure 2] and key concerns [Figure 3].

Figure 2. Inductive coding of key submission themes
Cynthia Khoo submitted these responses to many different ministries, again we have yet to hear a response.

Figure 3. Automated tone analysis of submissions
Creating a dedicated institution to handle and coordinate these issues — like Tiger Teams in the past — is an essential step to ensure the social impacts of AI are given necessary attention on an ongoing basis.
With the time I have left, I have two more comments:
2. Close the loopholes for third-parties processing political data and ensure that Canadian political parties have meaningful rules about their use of AI.
There is clear consensus for solutions. A recent national IPSOS poll found “80% of Canadians say political parties should follow the same privacy rules as businesses and public institutions” see Figure 4.
While we focus on deepfakes and AI-assisted messages, we miss major changes in Canada’s political infrastructure. AI is changing how political parties and political actors process and analyze their data. National Observer reporter Rory White followed the Toronto election where then-mayoral candidate Brad Bradford transferred voter data to an international third-party to encourage his supporters to post AI-generated messages of support on his social media posts. This is an example of what we’re missing when we look onl
Nothing is more disingenuous than political parties wringing their hands about the threats of AI to democracy without the willingness to undertake basic reforms to the privacy law and data protection to ensure that they’re accountable for the ways they’re processing voter and personal information.
Bill C-36 does not address these uses nor the Election Act.

Figure 4. Public opinions on privacy regulations for political parties
3. Include Public AI and AI Commons as critical components of Canada’s AI strategy
Finally, I want to emphasize that key data commons — Wikipedia, Creative Commons, and the Internet Archive — as well as public service media are suffering. Bots and AI scraping dramatically raises the costs to provide free and public services online, see Figure 5. AI undermines the reciprocity necessary to keep these sites alive, creating a paradox when the content is used more and more with fewer and fewer maintainers.

Figure 5. Monthly page visits by agent type Source: Wikimedia Statistics
This is an exciting time for innovation and change if we recognize that Canada’s commitments to communication rights need to be renewed for an age of general AI. I am here today, having stepped out of an international symposium on data commons, and I can attest that there is a global dialog and Canada needs to better integrate these discussions into its federal AI strategies. Canada’s AI strategy needs to invest in public service media and open Internet institutions and global cooperation to help mitigate AI’s impacts.
Thank you.