AI-Assisted Triage for Work Permit Eligibility

Ask about this system

Questions go to the Helpful Places, not the vendor.

Your rights

• Right to Be Informed of AI UseApplicants are informed that an automated system is used during the processing of their IEC work permit application. A plain language notice is posted through all service delivery channels (internet, in person, mail, and telephone), the privacy notice on application forms has been updated, and a web notice was published on the day the system launched. This is required under the Directive on Automated Decision-Making.
• Right to a Human ReviewIRCC officers make the final decision to approve or refuse on all IEC work permit applications. Applications that receive an automated positive eligibility assessment are still manually assessed by officers for admissibility before a final decision is rendered. All applications not triaged for automated positive assessment receive a full individual review of both eligibility and admissibility by an officer. The system enables human override of its assessments, and overrides are logged.
• Right to ContestA recourse process is established for clients who wish to challenge a decision made on their IEC work permit application. Explanations of decisions that result in denial must be provided in plain language, including the role of the automated system, the criteria used, the output produced, and the principal factors that led to the decision. Clients are informed of relevant recourse options. Contact IRCC through official immigration service channels to initiate a review.
• Right to an Explanation of a DecisionWhere a decision results in the denial of a benefit or service, applicants have the right to a meaningful explanation in plain language that covers: the role of the automated system in the decision-making process; the training and client data used and how it was collected; the criteria used to evaluate their data; the system's output and how to interpret it; and the principal factors that led to the final decision. This explanation must also describe relevant recourse options. These requirements are mandated under the Directive on Automated Decision-Making (TBS).
• Right to Non-discriminationApplicants have the right to non-discriminatory treatment by this system. The model's rules are based only on data elements with a clear link to legislative, regulatory, and contractual requirements — not on protected characteristics. A Gender-Based Analysis Plus (GBA+) was conducted addressing impacts on gender and other identity factors. Rules are reviewed by officers, legal, policy, and data science experts for bias before introduction and at regular intervals. An ongoing quality assurance process monitors for discriminatory outcomes.

Risks and safeguards

• Civil liberties harmThe triage function could influence officer decision-making or introduce discriminatory outcomes in immigration processing.Safeguard: The system is limited to flagging clearly eligible applications for positive assessment and never refuses or recommends refusals; all admissibility decisions remain with officers. Rules are reviewed by experienced officers, legal, policy, and data science experts, and senior executives before introduction and at regular intervals. A Gender-Based Analysis Plus was conducted on the data. An ongoing quality assurance process monitors whether officers reach the same positive eligibility assessments as the system to detect any bias introduced. Triage output is presented neutrally and officers are trained never to let it substitute for their individualized assessment.
• Reputational harmApplicants in the non-triaged stream could experience longer processing times than those receiving automated positive assessments, creating unequal treatment perceptions.Safeguard: The model makes no recommendation regarding applications not triaged for automated positive eligibility assessments — each must be assessed individually by a decision maker. Clear messaging in officer training reinforces the obligation for individualized assessment. The project monitors for differential processing time impacts. The system cannot produce a stigmatizing label — it only facilitates approvals.