Thank you very much for the opportunity to be here today.
I'd like to begin by acknowledging that I'm speaking from unceded indigenous lands. The Kanien'kehà:ka nation is recognized as the custodians of the lands and waters on which I speak today.
My name is Fenwick McKelvey. I'm an associate professor in information and communication technology policy at the department of communication studies at Concordia University.
My remarks today draw on over 10 years of experience in algorithmic and AI governance in Canada. In that time, I've been a lead on major national and international research projects, a co-director of an AI research institute and a frequent commentator on Canada's AI strategy. I've also been a witness, critic and occasional participant in the pan-Canadian AI strategy.
I want to begin by emphasizing that Canada has a right to regulate AI. These technologies are inherently public technologies. In the same way that railways, highways and cellphones have required public regulation, so does AI.
AI regulation should not just be economic but social as well. The production and use of AI inherently interacts with human rights. The government has a duty to protect those rights, especially the growing gender-based impacts. Canada's branch plant industrial policy on AI, so far, has moved too slowly to protect these rights. We cannot focus on productivity while ignoring quality of work, or economic opportunity without seeing the winners and losers. Such myopia puts Canadians and governments at risk. Politicians and parties need to read the room and understand the growing public sentiment against the AI industry today.
AI regulation requires new approaches to commons management and protection of public knowledge systems. AI is a commons management issue. We are already feeling the strain generative AI has put on our collective institutions, like our media systems and our public knowledge infrastructures as well. We have a history of building public media in this country and also an opportunity to build public AI. We need to define a vision for the future of this technology that works for all Canadians. Future policy must ensure we have the regulatory capacity, independence and financial means to achieve this vision.
Universities are critical to steward the knowledge commons but are in crisis. The pan-Canadian AI strategy must extend beyond its three research centres and include greater participation from the social sciences, arts and humanities. If AI is truly a problem for all of Canadian society, that challenge demands the participation of all of Canada's public institutions and facilities from coast to coast.
Canada must do better with public participation when developing its AI policy. Sprints and AI-assisted “what we heard” reports undermine the ethos of democratic participation, contributing to consistent public opinion polls reporting low trust by Canadians around AI.
By contrast, a coalition of civil society organizations launched the People's Consultation on AI last year in French and English. The consultation received over 70 submissions. My fellow researcher Joanna Redden and I are working on analyzing these comments. At first glance, we were struck by the depth and the sophistication of the responses. To build Canada's capacity to regulate AI, we need to hear more from these voices from civil society and from those most affected by AI. They need better inclusion in how Canada listens and understands the impacts of this technology, both social and economic.
Indigenous voices need special consideration in Canada's AI governance. Canada's true potential with AI can only happen through reconciliation and respect for nation-to-nation governance that understands and celebrates the distinct knowledge systems contained within our shared territories along with their influence and significance for understanding these technologies we call AI.
The challenge ahead is daunting and is exacerbated by long-term regulatory gaps. We are now trying to deal with AI empires when in the past we have struggled with social media platforms and even newspaper barons. The task is great, but now is the time to act towards a more inclusive, social approach to AI governance that meets the challenges of our time.
Thank you very much.