How and why governments are regulating AI: a comparison of legislative frameworks

This paper examines different legislative approaches to regulate the production, deployment, and use of Artificial Intelligence (AI). Specifically, it compares how the European Union (EU), Canada, and the United States (US) craft their legislation to better understand the underlying policy goals that these governments seek to advance. The first part of the paper contextualizes the discussion by outlining the well documented dangers posed by unregulated AI systems as well as the technological narratives that have been constructed to craft public expectations of both government and industry. The second part of the paper explains the comparative exercise this paper undertakes and provides some procedural history on the current state of the laws being examined. The third section compares the approach of each piece of legislation in terms of scope, obligations, and compliance, while the final section discusses these findings. The paper notes that while the EU has established a comprehensive regulatory framework that prioritizes the protections of harms over other interests, the approach developed by both the US and Canada displays a reluctance to move away from existing self-regulatory models in preference of advancing commercial interests.