Canada Launches First Public Registry Tracking Government AI Use
Updated 2:20 AM GMT+8, Wed December 17, 2025
Canada has launched a public registry that lists how artificial intelligence is used across federal departments. The goal: bring clarity, oversight, and shared visibility to AI systems supporting operations and service delivery.
What the registry includes
- More than 400 AI projects and systems across departments.
- Coverage from early-stage research to tools already supporting services.
- Details on each system's purpose, description, intended or current use, and sourcing (internal build vs. external vendor).
- A single view to see where AI is active, planned, or piloted across government.
Why this matters for departments
This registry gives teams a shared map of AI activity. Use it to spot overlap, reduce duplicate builds, plan jointly with peers, and find proven solutions before starting from scratch.
How to use it right now
- Scan for use cases similar to your mandate; note what's live vs. in pilot.
- Map entries to your next 12-18 months of work and flag collaboration opportunities.
- Reach out to listed program owners to compare results, risks, and procurement paths.
- Align your procurement with what's already approved or evaluated elsewhere.
- Update your internal AI inventory so it matches the registry's fields.
- Set risk triggers (e.g., citizen impact, automation level, data sensitivity) and require review when thresholds are met.
Governance and risk basics to bake in
Anchor your processes to existing federal guidance. Use the Directive on Automated Decision-Making and the Algorithmic Impact Assessment to set risk tiers, documentation, testing, and transparency requirements.
- Run an impact assessment before production and after major model changes.
- Formalize vendor due diligence: model provenance, data sources, evaluation results, and change logs.
- Document privacy, security, accessibility, and bias testing in plain language.
- Publish user-facing explanations for any AI influencing decisions or service pathways.
Reference: Directive on Automated Decision-Making and Algorithmic Impact Assessment.
Prepare for 2026 public consultations
Public consultations will refine the registry's design and usability in 2026. Start gathering insights now so your input is specific and actionable.
- Collect feedback from program, procurement, privacy, security, and service teams.
- List missing fields (e.g., model type, data sources, evaluation metrics, human-in-the-loop steps).
- Propose usability improvements: filters, tags, API access, bulk updates, and cross-department contacts.
- Draft a short submission with examples where the registry helped or created friction.
Quick checklist for program and policy leads
- Assign an owner to keep your entries current and accurate.
- Audit all AI systems and pilots; ensure each is listed with clear purpose and status.
- Add links to plain-language explanations and service impacts where applicable.
- Align contracts and statements of work with disclosure, testing, and monitoring requirements.
- Set evaluation cadences (pre-launch, quarterly, post-model update) and record outcomes.
- Create a contact point for cross-department collaboration and knowledge sharing.
The registry makes AI work visible. Use that visibility to improve services, spend smarter, and build trust with citizens and stakeholders.
If your team needs skills development to support responsible AI delivery, explore curated options by role here: Complete AI Training - Courses by Job.
Your membership also unlocks:
