AI-Assisted Triage and Approval of Work Permit Applications

Eligibility & Public Benefits

What it collects that can identify you

Sensitive personal information

Identifiable data

Application data submitted directly by clients, including personal identifying information. Data is classified at Protected B / Protected C level and is held in IRCC's Global Case Management System (GCMS). Personal Information Bank PPU 068 governs this data.

AIA — About the Data (Q44–Q56)

Sensitive personal information

Identifiable data

Medical information from panel physicians (immigration medical exams), enforcement records from Canada Border Services Agency (CBSA), and information from Canadian or foreign law enforcement agencies. IRCC also obtains identity and background information from the US, Australia, and New Zealand.

AIA — About the Data (Q56)

Also collects operational data, which is anonymized data.

Run by: Immigration, Refugees and Citizenship Canada (IRCC)
Where: No fixed location
Kept: Retained As per Personal Information Bank PPU 068 and applicable legislation governing immigration records
Shared with: Accountable organization
Your copy: You cannot see the data it holds about you. What you can do

What it is for

Eligibility & Public Benefits

This system automatically triages in-Canada work permit applications by complexity and can issue positive eligibility determinations for routine post-graduate and other work permit applications on behalf of IRCC officers. It cannot refuse any application — only officers make final decisions, including all admissibility assessments. Applicants should know that automated tools may process their application before an officer reviews it.

What it collects and what happens to it

Data taken in

Sensitive personal information

Identifiable data

Application data submitted directly by clients, including personal identifying information. Data is classified at Protected B / Protected C level and is held in IRCC's Global Case Management System (GCMS). Personal Information Bank PPU 068 governs this data.

AIA — About the Data (Q44–Q56)

Sensitive personal information

Identifiable data

Medical information from panel physicians (immigration medical exams), enforcement records from Canada Border Services Agency (CBSA), and information from Canadian or foreign law enforcement agencies. IRCC also obtains identity and background information from the US, Australia, and New Zealand.

AIA — About the Data (Q56)

Operational data

Anonymized data

Application-level data from the Global Case Management System (GCMS), including program eligibility criteria, supporting documents, and multi-source records used to assess completeness and quality of each submission.

AIA — About the Decision (Q27) and About the Data (Q53, Q56)

Processing

Classification & Prediction

Rules-based classification assigns each application to a complexity tier (routine vs. complex) and, for routine applications, determines positive eligibility. No machine learning or statistical models are used — the rules are authored by IRCC staff based on program eligibility and admissibility criteria.

AIA — About the Algorithm (Q18–Q19) and About the Decision (Q24)

What it does

Deciding (Analytical AI)

Human decides

Rules-based triage classifies each work permit application into complexity tiers and scores routine applications as positively eligible. For complex applications, the tool assigns them to officers without making any eligibility determination. No machine learning is used — the system applies rules authored by IRCC staff.

AIA — About the Decision (Q23, Q24)

Outputs

A decision about you

Identifiable data

For routine applications, the system sends a positive eligibility determination directly to the Global Case Management System (GCMS). For all applications, an Excel spreadsheet showing the triage result and key application information is shared with processing officers. Officers make all final decisions including admissibility; the system cannot refuse or recommend refusal.

AIA — About the Decision (Q27)

A recommendation or prediction

Identifiable data

For complex applications, the tool triages the application to the appropriate officer tier without making an eligibility determination, providing a complexity classification and key application data in the Excel output to support officer decision-making.

AIA — About the Decision (Q20, Q24, Q27)

Run by

Immigration, Refugees and Citizenship Canada (IRCC)

The federal department that deploys and operates these automated triage tools for processing in-Canada work permit applications. Officers within IRCC make all final decisions on applications.

Government of Canada AI Register — 2526-IRCC-014

Built by

Not stated by the Helpful Places.

Kept for

Retained As per Personal Information Bank PPU 068 and applicable legislation governing immigration records

Application data and audit trails — including all system decisions and officer overrides — are retained in the Global Case Management System (GCMS) and system logs. Personal Information Bank PPU 068 governs retention. The system maintains audit trails that record all recommendations or decisions made, key decision points, and version information.
Duration: As per Personal Information Bank PPU 068 and applicable legislation governing immigration records

AIA — About the Data (Q48) and Procedural Fairness (Q18–Q25)

Shared with

Not available to me

Triage results and the Excel spreadsheets showing tier assignments cannot be disclosed to the public or to individual applicants, in order to safeguard the integrity of the immigration system. The specific rules used by the tools are also not publicly available.

AIA — About the Decision (Q27) and Project Description

Available to the accountable organization

IRCC officers access triage results and key application data via Excel spreadsheets to inform their processing decisions. The audit trail and system logs are also available internally to authorized IRCC staff for quality assurance and oversight purposes.

AIA — About the Decision (Q27) and Procedural Fairness Mitigations

Stored

Stored locally

Data is stored in Canadian federal government systems, primarily the Global Case Management System (GCMS). The data is controlled by the federal government and stored at Protected B / Protected C classification within Canada's federal IT infrastructure.
Duration: As per Personal Information Bank PPU 068 and applicable legislation

AIA — About the Data (Q49, Q50, Q53)

How to read the colours

Can it identify you?

Anonymized data: Data about people with the link to who is broken. Stripped of identifiers, blurred, aggregated, or noised so this system can’t reasonably tie a record back to an individual.
Pseudonymous data: Each person’s data is tied to a token (hash, ID, template) that lets this system recognise the same person across events, but the token itself doesn’t reveal a name. Reidentification is possible with extra information.
Identifiable data: The data either contains a direct identifier (name, address, account name, recognisable face or voice, plate number) or carries a token this system uses to look up legal identity during processing.

Who completes the loop?

Human decides: This mode suggests; a person decides what to do next. The AI is always advisory — a human is in the loop on every decision. Example: a triage tool ranks cases for a clinician who chooses which to see first.
Human executes: This mode decides; a person carries out the result. Example: an optimizer plans the day’s trash-collection routes, and drivers run them.
Autonomous: This mode decides and acts on its own. No person reviews each decision or carries out the resulting action.

Definitions from the DTPR standard. Amber is about your data, violet about who decides. The fuller the shape and the deeper the colour, the more identifying the data or the less a person is involved.

What you can do

Ask about this system

Questions go to the Helpful Places, not the vendor.

Your rights

Right to Be Informed of AI UseApplicants have the right to be informed that automated tools are used in the processing of their work permit application. A plain language notice was posted on the IRCC website on the day the tools were launched, and the privacy notice on application forms is updated to inform clients of these tools. All service delivery channels (internet, in person, mail, telephone) must provide this notice.
Right to an Explanation of a DecisionApplicants who receive a denial of a benefit or service are entitled to a meaningful plain-language explanation of the role of the automated system in the decision, the training and client data used, the criteria applied, the system output, and the principal factors leading to the decision. Explanations must also include information about recourse options. A general description is available through the Algorithmic Impact Assessment and IRCC's departmental website.
Right to a Human ReviewAll applications — including those that receive an automated positive eligibility determination — continue to receive an individualized assessment by an officer who makes the final decision. Officers are required to review all relevant file information and cannot be overridden by the tool. Applicants whose application is refused have 90 days to re-apply and have access to judicial review at the Federal Court.
Right to Non-discriminationApplicants have the right to non-discriminatory processing. The tools' rules are vetted for potential bias, and a Gender-Based Analysis Plus (GBA+) of the data and triage results is conducted. An ongoing quality assurance process monitors whether officers make the same positive eligibility determinations as the tools to ensure that biases have not been introduced. Rules are reviewed by experienced officers, legal, policy, and data science experts to ensure they are non-discriminatory and aligned with established eligibility criteria.

Risks and safeguards

Civil liberties harmAutomated processing of immigration applications could infringe on applicants' rights to individualized assessment and due process, and automation bias could unduly influence officer decisions. Mitigations: The tools only automate positive eligibility determinations and cannot refuse any application. Officers receive training to not let triage results influence their decisions. The user manual explicitly states triage bins are administrative groupings with no guidance on client risk. An ongoing quality assurance process monitors officer alignment with tool outputs. Rules undergo extensive review by officers, legal, policy, and data science experts before deployment and at regular intervals. A GBA Plus assessment is conducted. Plain-language notice is posted through all service delivery channels.
Reputational harmIncorrect triage or erroneous positive eligibility determinations could stigmatize or unfairly categorize applicants. The system operates in an area of intense public scrutiny. Mitigations: Rules are vetted for relevance, reasonableness, and potential bias before deployment and at regular intervals by experienced officers, legal, policy, and data science experts. A GBA Plus assessment is conducted on the data and triage results. An audit trail records all decisions and overrides. The system only automates positive determinations — it cannot make or recommend any negative determination.