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 Translation Drafting for Government Content

Translation & Language Access

What it collects

Operational data
Anonymized data
  • Source-language text from Statistics Canada documents — publications, reports, metadata, and any Microsoft Office file (such as Word) submitted as a translation request by a GC employee. No personal information is involved.
Run by
Statistics Canada (StatCan)
Where
No fixed location
Kept
Not stated by the Helpful Places.
Shared with
Accountable organization
Your copy
You cannot see the data it holds about you. What you can do

What it is for

Translation & Language Access

This system uses a fine-tuned machine translation model to generate initial translation drafts of Statistics Canada documents from English to French and vice versa. The drafts are reviewed and refined by human language experts and quality-check officers before use. The system is being integrated into existing computer-assisted translation tools and does not process personal information.

What it collects and what happens to it

Data taken in

Operational data
Anonymized data
  • Source-language text from Statistics Canada documents — publications, reports, metadata, and any Microsoft Office file (such as Word) submitted as a translation request by a GC employee. No personal information is involved.

Processing

Language Models
  • A neural machine translation model fine-tuned on top of Microsoft Azure Custom Translator, specialised on Statistics Canada bilingual products and metadata to handle domain-specific vocabulary and formatting requirements.

What it does

Creating (Generative AI)
Human decides
  • The model generates new translated text from source-language input. Human language experts and quality-check officers review every draft before it is accepted or used.
Understanding (Semantic AI)
Human decides
  • The neural translation model understands and encodes the meaning of source-language text before generating the target-language output, including statistical and domain-specific terminology from StatCan content.

Outputs

Generated content
Anonymized data
  • Initial translation drafts in the target official language (English or French), delivered inside existing CAT tools. The output preserves the input document format (e.g. Word). Drafts are not published directly — they are reviewed and edited by human translators.

Run by

Statistics Canada (StatCan)
  • Statistics Canada is the federal department responsible for this system. It is deploying the fine-tuned machine translation tool for use by Government of Canada employees processing translation requests.

Government of Canada AI Register — 2526-StatCan-020

Built by

Government of Canada
  • The system was developed by the Government of Canada by fine-tuning on top of Azure Custom Translator, a Microsoft cloud service. No external vendor is separately identified as the technology provider.

Government of Canada AI Register — 2526-StatCan-020

Kept for

Not stated by the Helpful Places.

Shared with

Available to the accountable organization
  • Translation drafts are accessible to Statistics Canada language experts, quality-check officers, and the GC employees who submitted the translation requests, via integration with existing CAT tools.
Not available to me
  • The system is an internal government tool for GC employees. Members of the public do not have access to the translation drafts or the system interface.

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 UseThe register confirms that AI use is disclosed to users. GC employees who submit translation requests are informed that drafts are produced by an AI system before human review.
  • Right to a Human ReviewAll AI-generated translation drafts are reviewed and corrected by human language experts and quality-check officers before any translation is finalised. No AI output is accepted without human oversight.

Risks and safeguards

  • Societal & cultural harmmachine translation errors or domain-specific mistranslations could propagate into official Statistics Canada publications, eroding bilingual information quality.Safeguard: the system produces drafts only; mandatory human expert review by language professionals and quality-check officers is required before any output is published or used officially.