AI Risk Scoring for Export Shipment Compliance

Border & Immigration · Enforcement · Risk Assessment & Triage

What it collects that can identify you

Sensitive personal information

Identifiable data

Export declaration data includes business numbers associated with exporters and shippers, which can identify legal entities and in some cases individuals. Classified as Protected A by the CBSA.

Algorithmic Impact Assessment — Export Compliance Dashboard

Also collects operational data, which is anonymized data.

Run by: Canada Border Services Agency (CBSA)
Where: No fixed location
Kept: Retained Not specified; subject to CBSA PIB CBSA PPU 1103 and applicable Government of Canada records 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

Enforcement

Risk Assessment & Triage

This system uses machine learning to assign a risk probability score to outgoing export shipments, helping Canada Border Services Agency officers decide which shipments to inspect for potential non-compliance. Officers use the score alongside their own knowledge and other systems — it does not make decisions on its own. The system involves personal information (such as business numbers and shipment data) but individuals are not currently informed that AI is being used.

What it collects and what happens to it

Data taken in

Sensitive personal information

Identifiable data

Export declaration data includes business numbers associated with exporters and shippers, which can identify legal entities and in some cases individuals. Classified as Protected A by the CBSA.

Algorithmic Impact Assessment — Export Compliance Dashboard

Operational data

Anonymized data

Shipment-level export declaration data including container numbers, HS codes, commodity descriptors, country of destination, and declared value — sourced from the CBSA Enterprise Data Warehouse (originally from CERS export declarations and SEE examination results).

Algorithmic Impact Assessment — Export Compliance Dashboard

Processing

Classification & Prediction

A machine learning statistical model that predicts the probability of non-compliance for each export shipment. The model is trained on historical export declaration and examination result data from the CBSA Enterprise Data Warehouse.

Algorithmic Impact Assessment — Export Compliance Dashboard

Anomaly Detection

In addition to risk scoring, the system identifies when specific data points in the export declaration are missing or anomalous relative to historical patterns.

Algorithmic Impact Assessment — Export Compliance Dashboard

What it does

Deciding (Analytical AI)

Human decides

The system classifies and scores export shipments by risk probability using a statistical/machine learning model. A CBSA Officer reviews the output and makes the final decision on whether to examine a shipment — the tool is explicitly advisory.

Algorithmic Impact Assessment — Export Compliance Dashboard

Outputs

A recommendation or prediction

Anonymized data

The system outputs a ranked list of export shipments with an associated risk probability score per shipment, and flags missing or anomalous data fields. This is advisory only — CBSA Officers make the final decision on examination referrals.

Algorithmic Impact Assessment — Export Compliance Dashboard

Run by

Canada Border Services Agency (CBSA)

The Canada Border Services Agency (CBSA) is the federal institution that developed and deploys this machine learning tool within its Commercial and Trade branch to support export compliance screening by CBSA Officers.

Government of Canada AI Register — Export Compliance Dashboard

Built by

Government of Canada

The machine learning model was developed in-house by the Government of Canada, specifically by the Canada Border Services Agency. No external vendor is involved in building the model.

Government of Canada AI Register — Export Compliance Dashboard

Kept for

Retained Not specified; subject to CBSA PIB CBSA PPU 1103 and applicable Government of Canada records retention schedules

The AIA identifies PIB CBSA PPU 1103 as the relevant Personal Information Bank. A full Privacy Impact Assessment is planned but not yet completed. Specific retention periods are not stated in the AIA.
Duration: Not specified; subject to CBSA PIB CBSA PPU 1103 and applicable Government of Canada records retention schedules

Algorithmic Impact Assessment — Export Compliance Dashboard

Shared with

Available to the accountable organization

Output data (risk scores and shipment lists) is available to CBSA Officers (GC employees) as the primary users. Access is controlled by a grant/monitor/revoke permissions process within the CBSA.

Government of Canada AI Register — Export Compliance Dashboard

Not available to me

Exporters and shippers whose data is assessed by the system do not have access to their individual risk scores or to the criteria used by the system. The algorithm is a trade secret and criteria are not shared for security reasons.

Algorithmic Impact Assessment — Export Compliance Dashboard

Stored

Stored locally

Data is stored within CBSA systems (Enterprise Data Warehouse) under federal government control. The system operates within a closed environment — no internet connections — and data is controlled by the federal government.
Duration: Not specified; subject to Government of Canada records retention schedules

Algorithmic Impact Assessment — Export Compliance Dashboard

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 UseCurrently, individuals and exporters are NOT informed that an AI system is being used in the assessment of their shipments. The register confirms 'AI use disclosed to users: N.' The Directive on Automated Decision-Making and the AIA require that a meaningful explanation be published for common decision results via a departmental website.
Right to a Human ReviewAll decisions about whether to examine a shipment are made by a CBSA Officer — the system is explicitly advisory. Officers can cancel examination referrals using existing system functionality. A recourse process is established for clients who wish to challenge a decision.
Right to ContestA recourse process is established for clients who wish to challenge a decision resulting from this system. Examination referrals can also be cancelled by an Officer, allowing the exporter to proceed with the export of goods. Contact the Canada Border Services Agency for further information on recourse options.
Right to Algorithmic TransparencyThe AIA requires that a meaningful explanation of how the system works be published in plain language on a departmental website, covering the system's role, input data and sources, criteria used to evaluate data, outputs, and principal factors behind decisions. Note: the specific algorithm is treated as a trade secret and its criteria are not publicly disclosed.

Risks and safeguards

Civil liberties harmThe system involves personal information (business numbers) and operates in an area of public scrutiny (border management); individuals are not informed that AI is being used in their shipment assessment (AI use not disclosed to users).Safeguard: The tool is partial automation — a CBSA Officer makes all final decisions; human override is enabled; an audit trail records all recommendations and decision points; a recourse process exists for clients to challenge decisions; and a Privacy Impact Assessment is planned.
Reputational harmAn inaccurate risk score could result in more frequent examination of compliant exporters, potentially affecting their reputation or business operations. The algorithm is a trade secret, limiting external auditability.Safeguard: Officers retain discretion to cancel referrals; the system is intended to improve targeting accuracy; documented processes for bias testing and data quality are in place; and a change-control log is maintained for model updates.