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 Disinformation Detection for Foreign Policy Analysis

Safety & Security · Planning & Decision-making

What it collects that can identify you

Sensitive personal information
Identifiable data
  • Named entity recognition extracts the names of specific individuals mentioned in media sources including news articles, social media, and licensed content. The register confirms this system involves personal information.

Also collects about behaviour and operational data, which is anonymized data.

Run by
Global Affairs Canada (GAC)
Where
No fixed location
Kept
Not stated by the Helpful Places.
Shared with
Not stated by the Helpful Places.

What it is for

Safety & SecurityPlanning & Decision-making

This system uses IBM Watson Natural Language Understanding to automatically extract named entities (people, organizations, locations) from news articles, social media, and licensed media sources, then scores content for relevance to foreign policy and reputational risk. It helps Government of Canada analysts detect disinformation narratives and foreign interference patterns. The system involves personal information and has since been retired from active use.

What it collects and what happens to it

Data taken in

About behaviour
Anonymized data
  • Social media posts and activity (e.g., Twitter content, RSS feeds) are ingested as behavioural data reflecting publishing patterns and information spread, processed to identify disinformation narratives.
Sensitive personal information
Identifiable data
  • Named entity recognition extracts the names of specific individuals mentioned in media sources including news articles, social media, and licensed content. The register confirms this system involves personal information.
Operational data
Anonymized data
  • Traditional media sources (news articles), licensed content aggregators such as Factiva, and RSS feeds are ingested as operational content for analysis.

Processing

Language Models
  • IBM Watson Natural Language Understanding uses deep learning to extract meaning and metadata from unstructured text, including named entity recognition for people, organizations, and locations.
Classification & Prediction
  • The system applies scoring methods to classify and rank media content according to its relevance to foreign policy and reputational risk, and detects patterns indicative of disinformation or foreign interference.

What it does

Sensing (Perceptive AI)
Human decides
  • IBM Watson NLU performs named entity recognition on unstructured text, turning raw media content into structured extractions of people, organizations, and locations. Human analysts review and act on the results.
Deciding (Analytical AI)
Human decides
  • The system scores articles for relevance to foreign policy and reputational risk, and applies analytical methods to detect disinformation narratives and patterns. All outputs are advisory — analysts make final decisions.

Outputs

A recommendation or prediction
Anonymized data
  • The system produces relevance scores, detected disinformation narratives, and pattern analyses as advisory outputs for Government of Canada analysts. These are recommendations and analytical findings, not binding decisions.
Sensitive personal information
Identifiable data
  • Named entity extraction produces structured outputs identifying specific individuals mentioned in analysed media content, surfaced to analysts as part of the disinformation detection dashboard.

Run by

Global Affairs Canada (GAC)
  • The federal department responsible for Canada's international relations, foreign policy, and trade. Global Affairs Canada deployed this dashboard to support analysts in detecting disinformation and foreign interference.

Government of Canada AI Register — Disinformation Dashboard

Built by

Torusoft Inc. and IBM
  • Torusoft Inc. developed the dashboard and IBM provided the underlying Watson Natural Language Understanding AI platform used for named entity recognition and deep learning text analysis.

Government of Canada AI Register — Disinformation Dashboard

Kept for

Not stated by the Helpful Places.

Shared with

Not stated by the Helpful Places.

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 TransparencyThis system has been retired. When active, it was an internal analytical tool used by Government of Canada employees; it was not used to make binding decisions about members of the public. Information about how the system operated is available through the Government of Canada's AI Register at the source URL. Further inquiries may be directed to Global Affairs Canada.

Risks and safeguards

  • Civil liberties harmMonitoring and extracting information about named individuals from media sources at scale raises civil liberties risks related to surveillance, freedom of expression, and due process — particularly for individuals who may be misidentified as disinformation actors. The system was advisory-only (human analysts make final decisions), which provides partial mitigation. The system has since been retired, eliminating ongoing risk. No further mitigation details are documented in the register.
  • Reputational harmAutomated entity extraction and disinformation scoring risks falsely associating individuals or organizations with disinformation activity, causing reputational damage. The human-in-the-loop design (analysts review all outputs) partially mitigates this. The system is now retired. No specific accuracy thresholds, correction procedures, or challenge mechanisms are documented in the register.