Lessons from ArriveCAN: Access to Information and Justice during a Glitch

Abstract

In summer 2022, ArriveCAN, Canada's border app mandated during the COVID-19 pandemic for travelers entering the country, began sending certain users erroneous notifications to quarantine. On July 14, 2022, the federal government identified a glitch that was responsible for sending these erroneous notifications and patched it six days la¬ter. However, the federal government only publicly acknowledged a glitch was responsible for sending the erroneous orders four days after that — a full 10 days after it had become aware of the problem. During that time, 10,200 people received erroneous quarantine orders. These orders were not minor inconveniences. They were physical restraints on mobility enforced through the maximum penalties of the Quarantine Act. The issuance of mandatory quarantine orders by an app reliant on automated decision-making and artificial intelligence raised elevated concerns about the mandatory use of such technologies by the federal government. This article describes this episode from the perspective of a party seeking transparency and accountability of ArriveCAN's decision-making and highlights the interrelated access to information and justice concerns generated by the glitch and the federal government's response to it. Recommendations are discussed for moving forward in the context of governmental insistence on the use of mandatory data collection, retention, and use in automated decision-making and artificial intelligence systems.