AI-Assisted Processing of Privately Sponsored Refugee Applications

Border & Immigration

What it collects that can identify you

Sensitive personal information

Identifiable data

Personal information from PSR applicants (refugees and sponsors), classified at Protected B/C. This includes data elements with a clear link to legislative and regulatory requirements for the PSR program, collected from applications submitted to IRCC. The system uses personal information in a decision-making process that directly affects those individuals.

AIA — About the Data, Q26–Q30

Also collects operational data, which is anonymized data.

Run by: Immigration, Refugees and Citizenship Canada (IRCC)
Where: No fixed location
Kept: Retained Not specified in the AIA; governed by IRCC's Personal Information Banks and standard government record retention schedules
Shared with: Accountable organization
Your copy: You cannot see the data it holds about you. What you can do

What it is for

Border & Immigration

This system helps Immigration, Refugees and Citizenship Canada officers process privately sponsored refugee applications more efficiently. It automatically approves routine sponsorship portions of applications and sorts refugee files to appropriate officers based on their expertise and office capacity. Critically, officers retain final decision-making authority on all applications — the system never refuses an application, and all refusals remain the sole responsibility of human officers.

What it collects and what happens to it

Data taken in

Sensitive personal information

Identifiable data

Personal information from PSR applicants (refugees and sponsors), classified at Protected B/C. This includes data elements with a clear link to legislative and regulatory requirements for the PSR program, collected from applications submitted to IRCC. The system uses personal information in a decision-making process that directly affects those individuals.

AIA — About the Data, Q26–Q30

Operational data

Anonymized data

Data about migration office capabilities and officer expertise used to assign refugee applicant files to appropriate officers. This operational data supports the workload-routing function of the system and is drawn from multiple different sources within IRCC.

AIA — Project Description

Processing

Classification & Prediction

The system applies rules-based classification to determine which sponsorship applications qualify as routine cases for automatic approval. Rules use data elements with a clear link to legislative and regulatory requirements and were reviewed by experienced officers, legal, policy, data science, and privacy experts.

AIA — Q19 (Impact Assessment detail)

Optimization

The system sorts and assigns refugee applicant files to officers at migration offices overseas based on office capabilities and officer expertise, optimizing the allocation of workload across the immigration network to reduce differences in processing times across offices.

AIA — Project Description

What it does

Deciding (Analytical AI)

Autonomous

The system classifies sponsorship applications into 'routine cases that can be automatically approved' versus those requiring officer review, and sorts refugee files to the appropriate migration office and officer. Automatic approvals of the sponsorship portion are the only automated positive determinations — no refusals are automated.

AIA — About the Decision (Q13)

Deciding (Analytical AI)

Human decides

For the refugee applicant portion of all applications, the system performs triage and assignment to officers, but all eligibility and admissibility decisions are made by officers. Officers also retain final authority even where the sponsorship portion has been automatically approved.

AIA — Project Description and Q13

Outputs

A decision about you

Identifiable data

Automated positive approval determinations for the sponsorship portion of routine PSR applications. The system only renders approvals — never refusals. Even where the sponsorship portion is automatically approved, officers continue to make the final decision on each complete application.

AIA — Q13 and Q19

A recommendation or prediction

Identifiable data

Officer and office assignment routing for the refugee applicant portion of applications. The triage output assigns files to officers but does not produce a recommendation on the substantive refugee decision — that assessment remains fully individualized by each officer.

AIA — Project Description

Run by

Immigration, Refugees and Citizenship Canada (IRCC)

The federal department responsible for deploying and operating the automation system that processes privately sponsored refugee applications. IRCC officers retain final decision-making authority on all applications processed by the system.

Government of Canada AI Register — IRCC entry

Built by

Not stated by the Helpful Places.

Kept for

Retained Not specified in the AIA; governed by IRCC's Personal Information Banks and standard government record retention schedules

Personal information and audit trail data are retained in accordance with IRCC's existing Personal Information Banks (PIBs) and Privacy Impact Assessments (PIAs). The audit trail records all decisions, key decision points, system versions used, authorized decision-makers, and all system changes.
Duration: Not specified in the AIA; governed by IRCC's Personal Information Banks and standard government record retention schedules

AIA — Section 3.1 Q29; Section 3.2 Q17–Q24

Shared with

Available to the accountable organization

Output data (automated decisions, triage assignments, and audit trail records) is available to authorized IRCC officers and the accountable organization. Access permissions are controlled, monitored, and revocable. The federal government controls all data and the system operates within a government environment with no open internet connections.

AIA — Section 3.2 Q26, Q31, Q35

Not available to me

Applicants (refugee claimants and sponsors) do not have direct access to the system's outputs or the rules used by the system for triage or automated positive sponsorship determinations. Officers will not be aware of the rules used by the system, nor will they receive detailed information about the system's analysis. A plain language public notice is required but detailed system logic is not publicly available.

AIA — Q19

Stored

Stored locally

Data is controlled by the federal government and stored within Government of Canada infrastructure. The system does not connect to the internet or external systems, and no third-party cloud storage is indicated. Data is classified at Protected B/C level.
Duration: Not specified in the AIA; governed by IRCC's Personal Information Banks and standard government record retention schedules

AIA — Section 3.1 Q30–Q34

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 UseA plain language notice is required to be posted through all service delivery channels in use (Internet, in person, mail, or telephone) informing clients that automation is used in processing their applications. This requirement flows from Impact Level 2 compliance under the Directive on Automated Decision-Making.
Right to ContestA recourse process is established for clients who wish to challenge a decision. The AIA confirms a mechanism exists for clients to seek reconsideration. Additionally, for decisions that result in the denial of a benefit or service, a meaningful explanation must be provided as required under Impact Level 2 of the Directive on Automated Decision-Making.
Right to a Human ReviewAll applications where the sponsorship portion cannot be automatically approved receive a full individualized assessment by officers in accordance with standard practice. Even where the sponsorship portion is automatically approved, officers make the final decision on each complete application. Officers can override system decisions, and all overrides are logged.
Right to Algorithmic TransparencyUnder Impact Level 2 of the Directive on Automated Decision-Making, IRCC must ensure documentation on the design and functionality of the system is available. The system is not a trade secret and its logic is not considered difficult to interpret or explain. The system is able to produce reasons for its decisions when required and the audit trail supports generation of notifications.

Risks and safeguards

Civil liberties harmThe system processes immigration applications for vulnerable refugee populations, a domain of intense public scrutiny and potential rights impacts. The algorithm could introduce bias or discriminatory outcomes in determining which cases receive expedited processing. Mitigations: The system never refuses applications — all refusals are human decisions. System rules use only data elements with a clear link to legislative and regulatory requirements. A Gender Based Analysis Plus of the data was completed. Regular monitoring and quality assurance measures are in place to identify and mitigate bias or discrimination early. Civil society and the Office of the Privacy Commissioner were consulted.
Reputational harmErroneous automated approvals of ineligible sponsorship applications — or systematic triage errors — could misallocate resources and unfairly disadvantage applicants whose files are assigned to less suitable officers. Mitigations: System rules were reviewed by experienced officers, legal, policy, data science, and privacy experts, as well as senior decision-makers. A comprehensive audit trail records all recommendations and decisions made by the system, identifies key decision points, links them to relevant legislation, and logs all system changes. Human override of system decisions is enabled and logged.