evidoria

← Back to browse

Good practice Imported

Canada's Federal AI Register — the First National Inventory of Government AI, Built to Include the Tools That Process Citizens' Own Access Requests

Canada · Ottawa · See the Canada profile · See the Ottawa profile

Evidence: Descriptive / self-reported Top 35% 67/100 · Ask Evidence Copilot about this practice

On 28 November 2025 Canada published the federal government's first public AI Register, logging 400+ AI systems across 42 institutions — including the AI-assisted redaction tools departments now use to process citizens' own Access to Information requests.

400+
AI systems logged at launch (2025-11-28)
42
Institutions reporting into the register (2025-11-28)

Details

Maturity
Pilot
Promoter
Treasury Board of Canada Secretariat
Period
2023–2025
Keywords
access to information, government transparency, AI governance, records management

Context

On 28 November 2025, Treasury Board President Shafqat Ali launched the Government of Canada's first public AI Register, fulfilling a commitment in the federal public service's 2025-2027 AI Strategy and disclosing where and how AI is being explored, developed or deployed across government.

Objectives

The register aims to give citizens a standing public view of which AI systems their government uses, including the tools that departments now use to process Canadians' own Access to Information and Privacy (ATIP) requests.

Activities

At launch the register compiled self-reported input from 42 institutions covering more than 400 AI systems, including ATIPXpress, an AI-assisted case-management platform used across multiple federal departments that adds natural-language search, automated workflows and history-based redaction suggestions for ATIP analysts to accept or reject, and Agriculture and Agri-Food Canada's Document Detective tool, which helps ATIP teams find duplicate or near-duplicate records; the Treasury Board Secretariat also coordinates an Informal Network on AI and Automation in ATI so departments can compare notes.

Results

The register launched with more than 400 AI systems logged across 42 institutions, providing the first standing public mechanism, refreshed on an ongoing basis, for citizens to see which AI tools their government uses on their own information requests.

Conclusions

The register is explicitly described as a minimum viable product rather than a finished audit, since it depends entirely on institutions self-reporting their own AI use, and neither the launch materials nor independent reporting yet quantify how much AI-assisted redaction has actually shortened ATIP processing times.

Implementation

Indicative cost
Medium (€50k–€500k)
Time to results
Medium (1–3 years)
Staffing & skills
Treasury Board of Canada Secretariat (coordinating body, sponsored by Treasury Board President Shafqat Ali), 42 reporting federal institutions, each self-reporting their own AI systems, ATIP analysts using AI-assisted tools such as ATIPXpress and Document Detective in day-to-day casework

Conditions for success

  • A formal mandate under the federal public service's 2025-2027 AI Strategy
  • A standing interdepartmental coordination mechanism (the Informal Network on AI and Automation in ATI) for departments to compare practices
  • A self-reporting process that institutions actually populate and keep current

Common failure modes

  • The register relies on institutions self-reporting their own AI use rather than independent audit
  • No published quantification yet of how much AI-assisted redaction has shortened ATIP processing times or backlog

Commonly funded by

National / regional programmes Own resources / municipal budget

Indicative funding routes for practices of this type — always check each programme's current calls and eligibility rules.

Do you run this practice? Claim it — verified implementers get a public contact pathway and can propose corrections.

Data sources

Where this practice's information was retrieved from, and when.

Similar practices you may find useful