The Classification of AI Chatbots as High-Risk Systems in Legal Frameworks to Reduce their Risk to Human Safety

Abstract

This paper was originally written for Dr. Alys Avalos Rivera’s English 114W course Language and Purpose. The assignment asked students to create a persuasive piece on one of five available topics to raise awareness about an issue of public interest and propose a feasible solution, while also addressing the solution’s limitations. This paper focuses on the topic of human and AI relationships. The paper uses APA citation style.

AI chatbots are one of the defining instruments of human and AI relationships today. As these chatbots are increasingly integrated into various aspects of human society, ranging from business to medical to personal uses, the need to address the risks that these chatbots pose to society is also elevated. This paper argues that classifying AI chatbots as “high-risk systems” under Canada’s newly proposed Artificial Intelligence and Data Act (AIDA) would reduce the risks posed by the personalized nature of AI chatbots. While critics argue that the stricter regulations imposed by this Act and it’s internationally aligning structure could lead to Canada falling behind in the global AI race, this paper shows how AIDA is more likely to support Canadian AI innovation rather than stifle it.