Advanced analytics to help IRCC officers process visitor records

Ottawa, October 12, 2022—Immigration, Refugees and Citizenship Canada (IRCC) is improving how it processes visitor record applications by expanding the use of advanced analytics and other automated technology. While a new automated tool is used to help sort and process all visitor record applications, it never refuses or recommends refusing applications. IRCC officers will always make the final decision to approve or refuse applications.

The use of advanced analytics and automation is part of IRCC’s commitment to leveraging technology responsibly to build a stronger immigration system, for the benefit of all its clients. This tool helps accelerate IRCC’s work by identifying routine applications for streamlined processing. When the tool finds routine applications, it can decide that an applicant is eligible, and send the file to an officer to decide if the applicant remains admissible to Canada and make the final decision. To do this, the tool uses rules developed with IRCC officers or through machine learning based on data from previous IRCC files. The advanced analytics tool is reviewed routinely to make sure it is working as intended and that the results are consistent with applications that receive a full human review.

For cases that are not considered routine, the automated tool helps sort and assign cases, based on their level of complexity, to officers for a full individual assessment and decision, using only rules developed with IRCC officers. It also provides case notes to help summarize basic application information for the officer.

This new automated tool takes on most clerical and repetitive tasks related to sorting applications, which maximizes IRCC officers’ efforts assessing applications and making final decisions. The expected results are quicker decisions for clients and faster processing times overall for visitor record applications.

IRCC is continuing work to responsibly develop and use data-driven technologies in line with all privacy requirements. As part of its compliance with the Treasury Board Directive on Automated Decision-Making, an algorithmic impact assessment was completed to assess the analytical models used for visitor record applications. The assessment categorized the impact level of the system as moderate, and numerous measures have been put into place to mitigate possible risks. These measures include an extensive review process for potential bias impacts, privacy and security elements built into the design of the system, and the ability of officers to overturn an automated eligibility decision.