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-Assisted Analysis for Competition Enforcement

Enforcement · Research & Development

What it collects

Operational data
Anonymized data
  • Internal government operational data including case documents, legal texts, and public records related to competition investigations and market studies. The register states the system does not involve personal information.
Sensitive personal information
Anonymized data
  • Confidential business information and sensitive case-related documents submitted through consultations and market studies. While not personal information in the privacy-law sense, this data is Protected B classified and carries significant confidentiality obligations.
Run by
Innovation, Science and Economic Development Canada (ISED)
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

EnforcementResearch & Development

COMPASS is a secure, government-operated generative AI platform that helps investigators and analysts at Innovation, Science and Economic Development Canada review large volumes of sensitive documents related to competition law enforcement and market studies. It uses large language models to summarise documents, detect patterns, build searchable knowledge bases, and monitor competition-related developments from digital sources. The system processes internal and confidential information but is not reported to involve personal information. Government employees are informed that AI is in use.

What it collects and what happens to it

Data taken in

Operational data
Anonymized data
  • Internal government operational data including case documents, legal texts, and public records related to competition investigations and market studies. The register states the system does not involve personal information.
Sensitive personal information
Anonymized data
  • Confidential business information and sensitive case-related documents submitted through consultations and market studies. While not personal information in the privacy-law sense, this data is Protected B classified and carries significant confidentiality obligations.

Processing

Language Models
  • Large language models (LLMs) are the core processing technology, used for advanced text analysis, summarisation, and pattern detection across sensitive competition-related documents. The platform is Protected B compliant.
Search & Retrieval
  • Builds searchable knowledge bases from sensitive case documents and legal texts, enabling investigators to retrieve relevant information quickly from large document corpora.

What it does

Understanding (Semantic AI)
Human decides
  • Uses large language models to understand and search sensitive case documents, legal texts, and consultation inputs — building searchable knowledge bases and surfacing relevant patterns for human investigators and analysts to act upon.
Creating (Generative AI)
Human decides
  • Generates summaries, analytical outputs, and synthesised insights from large volumes of sensitive documents, enabling investigators to process information more efficiently. Human analysts review and act on these generated outputs.
Deciding (Analytical AI)
Human decides
  • Detects patterns and trends across large volumes of competition-related data, scoring and ranking insights to guide investigative priorities. Human investigators decide how to act on these analytical outputs.

Outputs

Generated content
Anonymized data
  • Summaries, analytical reports, and synthesised insights generated from competition case documents and market study inputs. Outputs are used internally by government employees and do not involve personal information.
A recommendation or prediction
Anonymized data
  • Pattern detections, trend identifications, and investigative insights surfaced to human investigators who then decide on enforcement or advocacy actions. The AI advises; human officials determine next steps.

Run by

Innovation, Science and Economic Development Canada (ISED)
  • The federal department that deploys and operates COMPASS to support competition enforcement and promotion activities, including investigations, legal analysis, and market studies.

COMPASS AI Register Entry

Built by

Government of Canada
  • The Government of Canada developed COMPASS internally. No third-party vendor is identified in the register entry.

COMPASS AI Register Entry

Kept for

Not stated by the Helpful Places.

Shared with

Available to the accountable organization
  • Outputs and processed data are available only to Government of Canada employees (primary users: GC employees). The system operates within a Protected B compliant secure environment.
Not available to me
  • Outputs are not accessible to the public, businesses, or other individuals. The system is an internal government tool for competition enforcement staff only.

Stored

Stored on 3rd Party Cloud
  • The register identifies this as a Protected B compliant generative AI platform. Government of Canada Protected B systems are typically hosted on approved cloud infrastructure meeting Treasury Board Secretariat security standards, though specific hosting details are not disclosed in the register entry.
  • Duration: Not specified in register
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 TransparencyThe Government of Canada discloses AI use in COMPASS on its public AI register. Parties to competition proceedings may have additional rights to understand how AI-assisted analysis informed enforcement decisions. For inquiries, contact Innovation, Science and Economic Development Canada.
  • Right to Be Informed of AI UseAI use is disclosed to users of the system (GC employees). The system's existence and purpose are disclosed on the Government of Canada's public AI register, accessible at the source URL above.

Risks and safeguards

  • Civil liberties harmConfidential business information and sensitive case data processed by an AI system could be exposed, misused, or used to reach conclusions that affect parties in enforcement proceedings without adequate procedural safeguards.Safeguard: The platform is Protected B compliant, limiting access to authorised GC employees only. Human investigators review AI outputs before any enforcement action is taken. AI use is disclosed to users.
  • Reputational harmAI-generated pattern detections or summaries could mischaracterise businesses or individuals subject to competition investigations, leading to unjust reputational damage if outputs are acted upon uncritically.Safeguard: The system supports human investigators who retain decision-making authority; outputs are analytical aids rather than binding determinations. The platform's secure environment limits exposure of preliminary findings.