AI-Assisted Legal Research for EI Adjudicators

Employment & Work · Inform

What it collects

Operational data

Anonymized data

Publicly available legal decisions from the Social Security Tribunal (SST), Federal Court of Appeals (FCA), Supreme Court of Canada (SCC), and Federal Court. These are institutional records, not personal information about claimants.

About behaviour

Anonymized data

Natural language queries entered by Benefit Officers describing the legal issue or fact pattern they need jurisprudence for. No personal information about claimants is included in these queries per the register disclosure.

Run by: Employment and Social Development Canada (ESDC)
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

Employment & Work

Inform

This tool helps Service Canada officers who review Employment Insurance claims find relevant court and tribunal decisions to support their adjudication work. It uses generative AI and retrieval-augmented generation to search legal databases including the Social Security Tribunal, Federal Court of Appeal, and Supreme Court of Canada. The tool is advisory — officers make the final decisions. It does not process personal information about EI claimants.

What it collects and what happens to it

Data taken in

Operational data

Anonymized data

Publicly available legal decisions from the Social Security Tribunal (SST), Federal Court of Appeals (FCA), Supreme Court of Canada (SCC), and Federal Court. These are institutional records, not personal information about claimants.

About behaviour

Anonymized data

Natural language queries entered by Benefit Officers describing the legal issue or fact pattern they need jurisprudence for. No personal information about claimants is included in these queries per the register disclosure.

Processing

Search & Retrieval

Retrieval-Augmented Generation (RAG) pipeline that indexes and searches the corpus of Canadian jurisprudence (SST, FCA, SCC, Federal Court decisions) in response to officer queries, feeding relevant passages into the generative model.

Language Models

Large language model used to understand officer queries in natural language and generate synthesized responses grounded in retrieved jurisprudence, as part of the Generative AI and NLP capability described in the register.

What it does

Understanding (Semantic AI)

Human decides

Natural Language Processing enables officers to query legal databases in plain language and retrieve semantically relevant jurisprudence. The system understands intent and surfaces matching decisions; the officer decides which references to use.

Creating (Generative AI)

Human decides

Generative AI capability produces summaries or synthesized responses to legal research queries, helping officers understand the relevant jurisprudence. All generated content is advisory and subject to officer review.

Deciding (Analytical AI)

Human decides

Retrieval-Augmented Generation (RAG) retrieves and ranks the most relevant legal decisions from the indexed corpus of SST, FCA, SCC, and Federal Court decisions to surface the best matches for a given query.

Outputs

A recommendation or prediction

Anonymized data

Relevant legal references (court and tribunal decisions) and AI-generated summaries surfaced to Benefit Officers to support their EI adjudication assessments. These are advisory outputs — the officer retains final decision authority.

Generated content

Anonymized data

AI-synthesized summaries and responses to legal research queries generated by the language model component, grounded in retrieved jurisprudence. Content is presented to officers as a research aid, not as authoritative legal advice.

Run by

Employment and Social Development Canada (ESDC)

Federal department responsible for deploying this AI tool to support Service Canada Benefit Officers conducting Level 2 Adjudications, Requests for Reconsiderations, and appeals of Employment Insurance claims.

GC AI Register — EI Jurisprudence Search Aid Tool

Built by

Accenture

Technology vendor contracted to develop the EI Jurisprudence Search Aid Tool on behalf of Employment and Social Development Canada.

GC AI Register — EI Jurisprudence Search Aid Tool

Kept for

Not stated by the Helpful Places.

Shared with

Available to the accountable organization

Outputs (legal references and AI-generated summaries) are available to Benefit Officers within Employment and Social Development Canada / Service Canada who use the tool in their adjudication work.

Not available to me

The tool is an internal government employee tool. Members of the public — including EI claimants whose appeals may be adjudicated by officers using this tool — do not have direct access to the system or its outputs.

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 TransparencyThe Government of Canada has disclosed this AI tool in the public Algorithmic Impact Assessment Register. The register entry describes the tool's capabilities (Generative AI, NLP, RAG), data sources, and the fact that it does not process personal information. Primary users are government employees (Benefit Officers), not members of the public. For general information, contact Employment and Social Development Canada.
Right to Be Informed of AI UseThe register confirms that AI use is disclosed to users of this tool. Benefit Officers who use the system are informed that they are working with an AI-assisted research tool. Members of the public whose EI claims may be informed by this tool's outputs are not direct users of the system.

Risks and safeguards

Reputational harmAI-generated legal summaries or retrieved decisions could be inaccurate, misapplied, or misattributed, potentially leading to incorrect EI adjudication outcomes that harm claimants.Safeguard: The tool is explicitly advisory — Benefit Officers retain final decision authority and are responsible for verifying references. The system surfaces supporting material rather than making binding determinations. AI use is disclosed to officer-users per the register.
Loss of autonomyOfficers may over-rely on AI-surfaced jurisprudence, reducing independent legal judgment and causing automation bias in adjudication decisions affecting EI claimants.Safeguard: The system is designed as a search aid, not a decision-maker. Officers retain full adjudication authority. AI use is disclosed, ensuring officers are aware they are using an AI tool and can apply professional judgment accordingly.