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 Assistant for Government Employees

Inform

What it collects that can identify you

About behaviour
Pseudonymous data
  • Text prompts and queries entered by GC employees during their interactions with the chatbot. These represent the employee's work-related queries, requests, and instructions.
Run by
Canadian Radio-television and Telecommunications Commission (CRTC)
Where
No fixed location
Kept
Not stated by the Helpful Places.
Shared with
Accountable organization

What it is for

Inform

CANChat is a generative AI chatbot available to Government of Canada employees, providing general-purpose assistance such as drafting, research, and information retrieval. It is built on a large language model and is designed to be used in a safe and secure environment. The system is a pilot program operated by the Canadian Radio-television and Telecommunications Commission (CRTC). Citizens are not direct users, but government decisions and communications produced with its help may indirectly affect them.

What it collects and what happens to it

Data taken in

About behaviour
Pseudonymous data
  • Text prompts and queries entered by GC employees during their interactions with the chatbot. These represent the employee's work-related queries, requests, and instructions.

Processing

Language Models
  • CANChat is built on a large language model (LLM) that processes employee text prompts and generates natural-language responses for general-purpose government work tasks.

What it does

Creating (Generative AI)
Human decides
  • CANChat generates text responses (drafts, summaries, answers) in response to employee prompts. A human employee reviews and decides whether to use any generated output.
Understanding (Semantic AI)
Human decides
  • CANChat uses language understanding to interpret employee queries and retrieve or synthesize relevant information, grounding responses in prior context within the session.

Outputs

Generated content
Pseudonymous data
  • Text responses, drafts, summaries, and answers generated by the LLM in response to employee prompts. Outputs are intended for internal government use and reviewed by the employee before any onward use.

Run by

Canadian Radio-television and Telecommunications Commission (CRTC)
  • The CRTC is the Government of Canada department that operates and deploys CANChat as a pilot generative AI tool for its employees.

Government of Canada AI Register — CANChat (2526-CRTC-004)

Built by

Government of Canada
  • The Government of Canada developed CANChat, making the federal government both developer and accountable body for this AI system.

Government of Canada AI Register — CANChat (2526-CRTC-004)

Kept for

Not stated by the Helpful Places.

Shared with

Available to the accountable organization
  • Interaction data and outputs are accessible to the CRTC and the Government of Canada as the operating and accountable organization. Access by external parties or the public is not indicated by the register.

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 UseGC employees using CANChat are informed that they are interacting with a generative AI system. The register describes CANChat as a publicly disclosed AI pilot, supporting the right of users to know an AI system is in operation.
  • Right to Algorithmic TransparencyThe Government of Canada publishes CANChat in its public AI register, disclosing the system's capabilities, developer, and primary users. Employees and citizens can consult the register to understand the system's general logic and scope.

Risks and safeguards

  • Psychological harmEmployees may over-rely on AI-generated content or be exposed to inaccurate or hallucinated outputs, leading to poor decisions or stress from correcting errors.Safeguard: The system is designed for use in a 'safe, secure and open way' as a pilot, implying human review of all outputs before use. Employees retain decision authority over all generated content.
  • Societal & cultural harmAI-generated government communications or documents that contain errors, biases, or lack cultural sensitivity could affect public trust or produce inequitable outcomes for citizens.Safeguard: As a pilot with human review of outputs, the risk is partially mitigated. The register does not disclose additional bias testing, fairness audits, or specific safeguards beyond the general 'safe and secure' framing.