Skip to content
This is NOT an official site of the Government of Canada. Click here for the official AI registry.

Helpful Places

Government of Canada – DTPR AI Register

Back to your search

AI-Assisted Document Classification for Government Records

Planning & Decision-making

What it collects that can identify you

Operational data
Identifiable data
  • Documents stored on Library and Archives Canada's Network Shared Drives (NSD). The register confirms the system involves personal information, so some of these documents may contain personal data about individuals. Approximately 10 million documents were in scope.
Run by
Library and Archives Canada (LAC)
Where
No fixed location
Kept
Not stated by the Helpful Places.
Shared with
Accountable organization

What it is for

Planning & Decision-making

This system automatically classifies approximately 10 million documents from Library and Archives Canada's internal network drives into GCdocs, the federal government's document management system. It also identifies and flags redundant, obsolete, and trivial content for elimination. The system involves personal information and Government of Canada employees were informed of the AI's use. As of the time of this record, the system has been retired.

What it collects and what happens to it

Data taken in

Operational data
Identifiable data
  • Documents stored on Library and Archives Canada's Network Shared Drives (NSD). The register confirms the system involves personal information, so some of these documents may contain personal data about individuals. Approximately 10 million documents were in scope.

Processing

Classification & Prediction
  • The system classifies documents into predefined categories aligned with the government's information architecture and predicts which documents constitute redundant, obsolete, or trivial (ROT) content. It also generates metadata labels to support downstream retrieval and management in GCdocs.

What it does

Deciding (Analytical AI)
Human decides
  • The system classifies documents and assigns them to categories or flags them for elimination. GC employees review and act on these classifications; the final records management decisions remain with human staff.
Understanding (Semantic AI)
Human decides
  • The system extracts meaning from document content to support classification and metadata creation, interpreting document topics, types, and relevance to assign appropriate categories within the government's information architecture.

Outputs

Operational data
Identifiable data
  • Classified documents with assigned metadata and information architecture tags, migrated into GCdocs. Also includes flags identifying redundant, obsolete, and trivial (ROT) content for potential elimination. Because source documents may contain personal information, outputs inheriting that content remain identifiable.

Run by

Library and Archives Canada (LAC)
  • Library and Archives Canada (LAC) is the federal institution that deployed this system to classify and migrate approximately 10 million documents from its internal network shared drives into GCdocs.

Government of Canada Algorithmic Impact Assessment Registry — 2526-LAC-BAC-007

Built by

Oproma
  • Oproma is the vendor that developed and supplied the automated document classification system for Library and Archives Canada, selected through a Request for Proposals (RFP) process.

Government of Canada Algorithmic Impact Assessment Registry — 2526-LAC-BAC-007

Kept for

Not stated by the Helpful Places.

Shared with

Available to the accountable organization
  • Classified documents and generated metadata are available to Library and Archives Canada and GC employees using GCdocs as the enterprise document management system.

Stored

Not stated by the Helpful Places.

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 UseGovernment of Canada employees have been informed that AI is used in this document classification process. The AI register entry confirms that AI use was disclosed to users.
  • Right to Algorithmic TransparencyThis system is listed on the Government of Canada's public AI register, providing transparency about its existence, purpose, data sources, and accountability. Employees can consult the register at open.canada.ca for general information about how the system operates.

Risks and safeguards

  • Reputational harmMisclassification risk: documents could be incorrectly categorized or incorrectly flagged as redundant, obsolete, or trivial, potentially leading to the inappropriate elimination of records containing personal information or records of legal or historical significance. The system was developed through an RFP process with a vendor, and GC employees are the primary users who review outputs. No specific mitigation measures are described in the register entry beyond human review by GC employees.
  • Civil liberties harmPrivacy risk: the system processes documents that may contain personal information about individuals. Automated classification and metadata creation could expose personal data to broader access within GCdocs than was previously the case on network shared drives. The register confirms personal information is involved but does not detail specific privacy safeguards beyond the general GCdocs access controls environment.