Is It Fair? A Case Study of Automated Decision Making in Canadian Migration Management

Advanced predictive analytics in the form of Artificial Intelligence or Machine Learning is becoming part of the arsenal employed by countries to manage their migratory flows. Canada applies these technologies to particular challenges such as improving process efficiency via Automated Decision Making. Inscribed in a trend that seems unstoppable, it is primordial that researchers investigate this new institutional practice, which has implications for privacy, social justice, equality, human rights, among others. This case study reconstructs the implementation of an existing Automated Decision Making system to explore many Algorithmic Fairness considerations. Immigration, Refugees, and Citizenship Canada is aware of the risks and deliberate in its adoption of these technologies. The Department follows a concerted approach with other actors within the Federal Government, which fundamentally lacks in meaningful transparency. Evaluating the adequate level of Algorithmic Fairness to expect from the Department is further complicated by its modulated agency and responsibility over applicants across its programs. Keywords: Migration Management, Automated Decision Making, Artificial Intelligence, Machine Learning, Algorithmic Fairness