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

AI-Generated Alt-Text for Census Images

Accessibility

What it collects

About a place
Anonymized data
  • Maps published in Census of Population articles — these depict geographic areas, boundaries, and statistical distributions across Canadian territory. No personal information is contained in these images.
About a measurement
Anonymized data
  • Charts, graphs, and video captures published in Census of Population articles, depicting aggregate statistical data (e.g. population counts, demographic distributions). These are aggregate-level images with no personal identifiers.
Run by
Statistics Canada (StatCan)
Where
No fixed location
Kept
Not stated by the Helpful Places.
Shared with
Accountable organization

What it is for

Accessibility

This tool automatically drafts alt-text descriptions for maps, charts, and other images published by Statistics Canada's Census of Population program. Government employees review and verify the AI-generated text before it is published, improving the accessibility of census content for people using screen readers or other assistive technologies. No personal information is collected or processed by this system.

What it collects and what happens to it

Data taken in

About a place
Anonymized data
  • Maps published in Census of Population articles — these depict geographic areas, boundaries, and statistical distributions across Canadian territory. No personal information is contained in these images.
About a measurement
Anonymized data
  • Charts, graphs, and video captures published in Census of Population articles, depicting aggregate statistical data (e.g. population counts, demographic distributions). These are aggregate-level images with no personal identifiers.

Processing

Computer Vision
  • The system interprets images — maps, charts, graphs, and video captures — extracting visual features and semantic content to support alt-text generation. This is the perceptive stage of the pipeline before text is generated.
Language Models
  • A language model generates bilingual (English and French) natural-language alt-text drafts from the visual features extracted by the computer vision stage.

What it does

Sensing (Perceptive AI)
Human decides
  • The system reads images (maps, charts, graphs, video captures) and extracts structured semantic content from them, producing a text description. A human employee then reviews and verifies the draft before it is published.
Creating (Generative AI)
Human decides
  • The system generates a new natural-language alt-text draft that did not exist before. This draft is produced in both English and French. Employees verify the generated text rather than writing from scratch.

Outputs

Generated content
Anonymized data
  • Bilingual (English and French) alt-text draft descriptions for census images, which Government of Canada employees review and verify before publication. The output contains no personal information.

Run by

Statistics Canada (StatCan)
  • Statistics Canada is the Government of Canada department responsible for producing data to help Canadians better understand their country. It developed and deploys this tool to auto-generate alt-text drafts for census images.

Government of Canada Algorithmic Impact Assessment Register — Census Alt-text Generator

Built by

Not stated by the Helpful Places.

Kept for

Not stated by the Helpful Places.

Shared with

Available to the accountable organization
  • The generated alt-text drafts are accessible to Government of Canada employees (primary users) who review and verify them before publication. The outputs are not available to the general public prior to editorial review.

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 Algorithmic TransparencyStatistics Canada has disclosed the use of AI in this system on the Government of Canada's AI Register. The system generates draft alt-text that is reviewed by employees before publication; the AI's role in producing accessibility content is publicly acknowledged.
  • Right to Be Informed of AI UseThe fact that AI is used in this process has been disclosed to users (Government of Canada employees who interact with the tool). The use of AI is confirmed as disclosed in the official register entry.

Risks and safeguards

No risks or safeguards have been published for this system yet.