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 Coding Assistant for Java Software Development

Employment & Work

What it collects

Operational data
Anonymized data
  • Java source code snippets from CRA internal systems are provided as input to the coding assistant at runtime. These are system-level code fragments, not personal data.
Run by
Canada Revenue Agency (CRA)
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

This tool uses generative AI to help Canada Revenue Agency software developers write and review Java code. It can detect anomalies in code, analyze text, and automatically generate code suggestions to reduce errors and improve consistency. It is intended for use by Government of Canada employees only and does not involve any personal information.

What it collects and what happens to it

Data taken in

Operational data
Anonymized data
  • Java source code snippets from CRA internal systems are provided as input to the coding assistant at runtime. These are system-level code fragments, not personal data.

Processing

Language Models
  • The system is built on open-source large language model technology, specialized or fine-tuned for Java code generation, text analysis, and anomaly detection within a software development context.
Anomaly Detection
  • Anomaly detection is applied to code snippets to identify unusual patterns, potential bugs, or deviations from consistent coding practices before developers commit changes.

What it does

Creating (Generative AI)
Human decides
  • The system generates Java code suggestions and completions in response to developer prompts. Developers review and decide which suggestions to accept, modify, or reject.
Understanding (Semantic AI)
Human decides
  • The system performs text and speech analysis to understand developer queries and code context, matching intent to appropriate code patterns or documentation.
Deciding (Analytical AI)
Human decides
  • The system performs anomaly detection on code snippets, flagging unusual patterns, potential errors, or deviations from coding standards for developer review.

Outputs

Generated content
Anonymized data
  • The system outputs AI-generated Java code suggestions, completions, and corrections. These are presented to developers as advisory recommendations, not applied automatically.
A recommendation or prediction
Anonymized data
  • The system produces anomaly flags and code quality recommendations, surfacing potential issues for developer review and decision. No automated action is taken without human approval.

Run by

Canada Revenue Agency (CRA)
  • The Canada Revenue Agency (CRA) is the federal department responsible for administering tax laws and benefit programs in Canada. CRA is deploying this generative AI coding assistant to support its software development employees.

Government of Canada AI Register — 2526-CRA-ARC-011

Built by

Not stated by the Helpful Places.

Kept for

Not stated by the Helpful Places.

Shared with

Not available to me
  • This system is for internal GC employee use only and does not process personal information about members of the public. Members of the public have no access to outputs from this system.
Available to the accountable organization
  • Code suggestions and outputs are available to CRA software development employees within the scope of their work. Access is limited to authorized GC employees.

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 Be Informed of AI UseAI use is disclosed to users of this system. GC employees using this tool are informed that they are interacting with an AI-powered coding assistant.
  • Right to Algorithmic TransparencyThis system has been publicly disclosed in the Government of Canada's AI and Algorithmic Impact Assessment register (entry 2526-CRA-ARC-011), which describes the system's capabilities, data sources, and status.

Risks and safeguards

  • Reputational harmAI-generated code suggestions could introduce subtle errors or vulnerabilities that reflect poorly on CRA systems if not properly reviewed.Safeguard: the system operates in a human-decides autonomy mode — all suggestions require developer review and approval before being adopted. The tool is positioned as advisory, not authoritative.
  • Societal & cultural harmOpen-source AI models used for code generation may reflect biases from training data, potentially enforcing non-inclusive or insecure coding patterns at scale across government systems.Safeguard: the system is currently in development status, with a more fit-for-purpose integrated tool being evaluated in parallel. Human developer review is required for all outputs.