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 Qualitative Analysis for Policy Evaluation

Planning & Decision-making · Research & Development

What it collects that can identify you

About behaviour
Identifiable data
  • Qualitative responses gathered from interviews and surveys — these are records of what people said or expressed, and may be identifiable depending on the research context and consent framework in place.

Also collects operational data, which is anonymized data.

Run by
Canadian Heritage (PCH)
Where
No fixed location
Kept
Not stated by the Helpful Places.
Shared with
Accountable organization, Vendor

What it is for

Planning & Decision-makingResearch & Development

This system explores the use of AI tools — including Microsoft Copilot and ChatGPT — to help Canadian Heritage employees summarize large amounts of qualitative information from interviews, surveys, and literature reviews, and to identify recurring themes and patterns. The system is currently in development and is intended only for use by Government of Canada employees, not the general public. Citizens should be aware that their interview or survey responses may be processed by AI tools to support departmental policy evaluation work.

What it collects and what happens to it

Data taken in

About behaviour
Identifiable data
  • Qualitative responses gathered from interviews and surveys — these are records of what people said or expressed, and may be identifiable depending on the research context and consent framework in place.
Operational data
Anonymized data
  • Literature reviews and published research documents — non-personal operational inputs that provide subject-matter context for the qualitative analysis.

Processing

Language Models
  • Large language models — specifically Microsoft Copilot and OpenAI ChatGPT — are being evaluated to process qualitative text, identify themes, and produce plain-language summaries.

What it does

Understanding (Semantic AI)
Human decides
  • The system uses semantic understanding to find recurring themes and patterns across qualitative text. Government of Canada employees review the outputs and make final analytical judgments.
Creating (Generative AI)
Human decides
  • The system uses generative AI capabilities to produce plain-language summaries and thematic syntheses of qualitative research data. Outputs are reviewed by GC employees before use.

Outputs

Generated content
Anonymized data
  • Plain-language summaries and thematic analyses derived from qualitative research inputs. Outputs are intended for internal GC employee use in policy evaluation and do not directly identify individuals.
A recommendation or prediction
Anonymized data
  • Identified recurring themes and patterns that guide policy evaluation findings. These are advisory outputs reviewed by GC employees, not binding determinations.

Run by

Canadian Heritage (PCH)
  • The federal department responsible for deploying and overseeing this AI-assisted qualitative analysis system for use by its employees in policy evaluation activities.

Government of Canada AI Register — 2526-PCH-002

Built by

Microsoft and OpenAI
  • Microsoft (Copilot) and OpenAI (ChatGPT) are the AI platform vendors whose tools are being evaluated for use in this system. They supply the underlying large language model capabilities.

Government of Canada AI Register — 2526-PCH-002

Kept for

Not stated by the Helpful Places.

Shared with

Available to the accountable organization
  • Outputs are available to Canadian Heritage employees for internal policy evaluation purposes. The register specifies GC employees as primary users.
Available to vendor
  • Data processed through Microsoft Copilot and OpenAI ChatGPT APIs may be accessible to these vendors depending on applicable service agreements and data-processing terms. This is not explicitly described in the register entry.

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 TransparencyInformation about this AI system is publicly available through the Government of Canada's Algorithmic Impact Assessment Register. Citizens whose interview or survey responses may be processed by this system should be informed of AI use through applicable research consent processes.
  • Right to a Human ReviewGC employees are identified as the primary users who review AI outputs before any analytical conclusions are reached. The system is advisory, with human review of all results prior to use in policy evaluation.

Risks and safeguards

  • Psychological harmInterview and survey participants whose qualitative responses are processed by AI tools may not have explicitly consented to AI-assisted analysis, potentially raising concerns about surveillance or misuse of sensitive disclosures.Safeguard: The system is in development and subject to ongoing evaluation; research consent frameworks should be updated to disclose AI use, and the scope of data processed should be limited to what is strictly necessary.
  • Societal & cultural harmUse of commercial LLMs (ChatGPT, Copilot) to synthesize qualitative research may introduce model biases that skew theme identification, potentially misrepresenting community perspectives in policy evaluation.Safeguard: The evaluation phase includes comparison of multiple tools; GC employees review all AI outputs before conclusions are drawn, providing a human check on AI-identified themes.