By: Maroussia Lévesque & Florian Martin-Bariteau (Center for International Governance Innovation)
In June of last year, the Canadian federal government proposed Bill C-27, which aimed to address the widespread use of digital technology. The bill included an unexpected element, the Artificial Intelligence and Data Act (AIDA). This act was designed to promote a responsible AI industry in Canada and to protect citizens from harmful AI systems through a risk-based, sanctions-backed approach. However, AIDA failed to achieve either objective.
AIDA, in fact, lacks substance and does not provide clear guidelines on the rights and obligations related to AI systems. The act contains undefined terms that hint at obligations for those responsible for AI systems, with the expectation that government regulations will fill in the gaps later. This approach, however, is subject to limited democratic oversight and creates a lack of legal certainty. Since critical provisions remain undetermined, AIDA is an incomplete legal framework.
