AI-Assisted Redaction for Access to Information Requests
Enforcement · Planning & Decision-making
What it collects that can identify you
- ATIP submissions from members of the public contain personal information about individuals, including requestor identity and personal details referenced in requested government records.
Also collects operational data, which is anonymized data.
- Run by
- Public Services and Procurement Canada (PSPC)
- 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
This system uses artificial intelligence to help government employees process Access to Information and Privacy (ATIP) requests more efficiently. It suggests which parts of documents should be redacted based on patterns and prior decisions, and uses natural language processing to assist with the redaction workflow. The AI makes suggestions that employees review and approve — it does not make final redaction decisions automatically. Members of the public who submit ATIP requests may have their submission documents processed by this system.
What it collects and what happens to it
Data taken in
- ATIP submissions from members of the public contain personal information about individuals, including requestor identity and personal details referenced in requested government records.
- GCdocs document content, ATIP database records, and historical ATIP documentation are used as operational inputs to train pattern recognition and support the redaction suggestion engine.
Processing
- Natural language processing models are used to read and interpret document text, identify sensitive patterns, and generate redaction suggestions based on content understanding and historical decisions.
- Custom and automated pattern reactions classify text segments as candidates for redaction based on recognized patterns, matching them to known categories of exemptions or exclusions under the Access to Information Act and Privacy Act.
What it does
- Natural language processing is used to understand document content, recognize patterns, and match text to redaction categories. Employees review and approve all suggested redactions before they are applied.
- History-based intelligent redaction suggestions are generated by analyzing patterns from prior ATIP decisions. The system scores and ranks candidate redactions for employee review, without making binding determinations.
Outputs
- The system outputs intelligent redaction suggestions — advisory annotations on document text — that government employees review and approve or reject. These are recommendations, not automatic redactions; a human makes the final determination.
Run by
- Federal department responsible for deploying the ATIPXpress application and its AI functionality to support the processing of Access to Information and Privacy requests across the Government of Canada.
Built by
- Technology vendor supplying the ATIPXpress application and its built-in AI capabilities, including natural language processing and intelligent redaction suggestions.
Kept for
Not stated by the Helpful Places.
Shared with
- Members of the public whose ATIP submissions are processed by this system do not have direct access to the AI-generated redaction suggestions or the internal processing outputs. Requestors receive only the final disclosed documents.
- ATIP analysts and authorized employees at Public Services and Procurement Canada have access to the AI-generated redaction suggestions within the ATIPXpress application as part of their review workflow.
Stored
- The ATIPXpress application and its AI-processed outputs are hosted by the vendor Opexus/Casepoint. The storage jurisdiction and specific cloud infrastructure details are not specified in the available register information.
- Duration: Not specified in available register information
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.
- AI registerGovernment of Canada Algorithmic Impact Assessment Registry — ATIPXpress AI Functionality (2526-PSPC-SPAC-001)Public Services and Procurement Canada, AI Register entry 2526-PSPC-SPAC-001.
- AI registerGovernment of Canada AI Register — 2526-PSPC-SPAC-001
- AI registerGovernment of Canada AI Register — 2526-PSPC-SPAC-001
- Register entryPublished by the Helpful Places. Reference 65466077. This disclosure was drafted with AI assistance.Schema: ai@2026-05-06-beta
What you can do
Ask about this system
Questions go to the Helpful Places, not the vendor.
Your rights
- Right to Be Informed of AI UseMembers of the public who submit ATIP requests should be informed that AI is used to assist in processing their submissions. Contact Public Services and Procurement Canada's ATIP Coordinator for information about how AI is used in processing your request.
- Right to a Human ReviewAll AI-generated redaction suggestions are reviewed and approved by a trained government employee before any redaction is applied to disclosed documents. No final redaction decision is made automatically without human oversight.
- Right to Algorithmic TransparencyYou have the right to know how this AI system works in general terms. The system uses natural language processing and historical redaction patterns to suggest portions of documents to redact. It does not make final decisions. Details are available through the Government of Canada's AI Register at open.canada.ca.
Risks and safeguards
- Civil liberties harmAI redaction suggestions applied systematically could suppress legitimate access to information by over-redacting documents, chilling public accountability.Safeguard: All suggestions require human approval before application; no automatic redaction occurs. The system is in development, providing time to establish accuracy benchmarks and review protocols before full deployment.
- Reputational harmPattern-based redaction suggestions could systematically misclassify text, leading to either unwarranted withholding or inadvertent disclosure of personal information in government records.Safeguard: Human review is required for all suggestions; historical documentation provides a training baseline; the system is being developed before full production deployment.