Proceed with Caution: Complicating Artificial Intelligence as a Future Decision-Maker in Canadian Law

Abstract

Artificial Intelligence [‘AI’] decision-making has been raised by practitioners, media, and even the Canadian government, as a solution to the many legal issues plaguing the legal system including costs, delays, and even bias. AI is increasingly being adapted by state organizations for decision-making purposes and in light of the COVID-19 pandemic, such use is anticipated to increase. However, I argue that in light of research discovering bias embedded in AI decision-making, such implementation is premature and could further deepen institutional inequalities. Furthermore, science is publicly perceived and branded by state actors as neutral and a marker of efficiency, thus identifying and eradicating such inequalities will be increasingly difficult as the layers of obscure AI decision-making and the legitimization of the neutrality of science compound. My paper uses recent changes in Canadian immigration law as a case study to demonstrate how a state can legitimize the use of technology with markers of efficiency and modernity while incorporating AI decision-making without transparency or in-depth consideration of the real risk and consequences of embedded bias. Immigration law is a fruitful example of the role that AI decision-making can play in entrenching inequality because it is a field of law that is directly engaged with maintaining borders and nation-building. Technology, and the manner in and purpose for which technology is used, reflects state values and state missions when it comes to race and society. My paper calls for a more cautious approach to AI decision-making that prioritizes transparency and accountability in order for AI decision-making to be a true investment in the future of the legal system.

Keywords: artificial intelligence, machine learning, race, bias, racism, technology, science technology studies, immigration, borders, administrative decision-making, nation-building, canada, fairness, transparency, nationhood