During a recent press conference in Vancouver, Canada’s Immigration Minister Sean Fraser highlighted the significant advancements achieved through the increased utilization of technology, advanced analytics, and streamlined processing, particularly focusing on temporary residency (TRV) applications for individuals awaiting permanent residency through family class sponsorship programs.
The transformative impact of these measures extends across all sectors of application processing within Immigration Refugees and Citizenship Canada (IRCC). As Canada prepares to welcome a record number of newcomers in the upcoming years, IRCC made over five million final decisions on applications in 2022, doubling the number from the previous year.
In a rapid shift towards the digitization and modernization of Canada’s immigration system, IRCC has embraced the increased use of Artificial Intelligence (AI) and advanced data analytics to enhance the speed of application processing. As a result, the streamlining efforts have led to an impressive 98% processing rate for spousal TRV applications, with new applicants experiencing significantly reduced processing times of just 30 days.
By digitizing its system and services, IRCC aims to better cater to the needs of clients and Canada as a whole, harnessing the power of new technological capabilities. This strategic utilization of technology enables more efficient application processing, allowing officers to dedicate their attention to handling complex cases.
IRCC incorporates artificial intelligence (AI), an advanced information technology that performs tasks typically carried out by humans, in various aspects of its processing, including:
- Automating positive eligibility determinations.
- Distributing applications among officers based on application characteristics.
- Identifying applications that may require additional verification.
- Managing workload distribution.
- Creating annotations that summarize basic client information to streamline searches in the Global Case Management System.
- Prioritizing client emails for faster responses and providing publicly available information to address inquiries.
- Assessing biometrics.
It’s important to note that none of the tools utilized by IRCC employ algorithms that autonomously accept or reject applicants. Decisions regarding visa or permit approvals are not solely based on computer-generated outcomes, ensuring that candidates are evaluated by human judgment.
What are the concerns?
There are concerns surrounding the use of AI by IRCC, primarily focusing on potential bias and a lack of transparency in decision-making during application processing.
For instance, the Treasury Board Directive on Automated Decision-Making, a government-wide policy, emphasizes that algorithms based on historical data can amplify existing inequalities related to race, class, gender, and other factors. It highlights the limitations of certain facial recognition software, which may not perform equally well across different skin colors or genders.
Relying too heavily on AI can also result in decision-making processes becoming less clear. According to the directive, the federal government must be capable of providing explanations for administrative decisions. Individuals who are denied services or benefits have the right to receive a reasonable and comprehensible explanation beyond simply attributing the decision to a computer.
Additionally, concerns have been raised regarding the use of “Chinook,” described by IRCC as a Microsoft Excel-based tool that simplifies the visual representation of client information. Chinook is utilized by IRCC officers to assess applications for temporary resident visas, study permits, and work permits. The department clarifies that Chinook does not employ artificial intelligence (AI) or advanced analytics for decision-making purposes, and there are no inherent decision-making algorithms integrated into the tool.
How does IRCC develop algorithms?
IRCC assures that it follows a Directive rooted in the Canadian Charter of Human Rights and Freedoms, which ensures equality rights and prevents discrimination. The department adheres to principles of transparency, accountability, legality, and procedural fairness, setting the guidelines for decision-making and the explanations provided to affected individuals.
Before implementing any new algorithm, IRCC undergoes an Algorithmic Impact Assessment (AIA). This mandatory risk assessment tool is part of the Treasury Board’s Directive on Automated Decision-Making. The AIA evaluates risk areas, mitigation strategies, and the potential impact of proposed algorithms. IRCC highlights that they were among the first government departments to utilize AIAs.
Furthermore, the department requires assessments of algorithm impacts, quality assurance measures for both data and algorithms, and proactive disclosure of how and where algorithms are employed.
To ensure alignment with the eligibility criteria outlined in the Immigration and Refugee Protection Act, experienced officers, legal experts, policy analysts, data scientists, privacy experts, and senior decision-makers regularly review the rules supporting the decision-making process.
Ongoing monitoring and quality assurance measures are in place to ensure the intended performance of these models and to identify and address any unforeseen adverse impacts at an early stage.
Minister Fraser maintains an optimistic outlook on the use of technology and advanced analytics by IRCC. He emphasizes that the technology adopted by the department enables a comprehensive evaluation of various factors in an individual’s application, aiding in determining their likelihood of eligibility for permanent residence and facilitating smoother processing by IRCC officers. He emphasizes that AI does not make final decisions.
“In the end, our officers still make every determination for eligibility. It is not possible for anyone to be denied solely through this technological solution.”