Responsible Use of Artificial Intelligence Technology - Office of the Chief Information Officer

Governance

Authority: Government of Newfoundland and Labrador

Audience: All employees, contractors, consultants, partners, students, volunteers, vendors, agents, third parties, and other persons working on behalf of the Government of Newfoundland and Labrador.

Issuing Public Body: Office of the Chief Information Officer

Original Issue Date: 2025-04-17
Date Last Reviewed: 2025-04-17

Version Number: 1.0

Notice:

This Policy follows the Government of Newfoundland and Labrador’s Accessible Communications Policy

This document is available in alternate format. Please contact OCIO@gov.nl.ca.

Forward questions and/or comments related to this document to OCIO@gov.nl.ca.

1.0 Overview
2.0 Application
3.0 Definitions and Acronyms
4.0 Statements
5.0 Roles and Responsibilities
6.0 Compliance and Enforcement
7.0 Supporting Materials and Version History

1.0 Overview

Artificial Intelligence (AI) technologies offer promise for improving the operations and services of the Government of Newfoundland and Labrador. The use of AI technology is guided by Government’s Guiding Principles in the Responsible Use of Artificial Intelligence and Policy – Responsible Use of Artificial Intelligence Technology.

AI is a rapidly evolving field that offers potential for innovation and efficiency of improved service but also poses risks that must be governed by policy and managed operationally.
This policy is established to enable the responsible use of AI within government departments, while ensuring appropriate governance. This policy includes an assessment process to evaluate AI technologies for responsible use.

The management and protection of information created and collected by government departments and other public bodies is subject to the requirements set out in the Management of Information Act (MOIA). The Office of the Chief Information Officer (OCIO) administers this Act, and in doing so, establishes policies, directives, standards, guidelines, and procedures required.

This policy will be reviewed annually, or as required by the Chief Information Officer.

^ Top of Page

2.0 Application

This policy applies to all government departments and (hereafter also referred to as “Government”) including all employees, contractors, consultants, partners, students, volunteers, vendors, agents, third parties and other persons working on behalf of the Government of Newfoundland and Labrador’s departments (hereafter referred to as “individuals”).

^ Top of Page

3.0 Definitions and Acronyms

Artificial Intelligence (AI) – Information technology that performs tasks that would ordinarily require biological brainpower to accomplish, (e.g., making sense of spoken language, learning behaviours, or solving problems). (Source: Government of Canada (GoC)).

AI Assessment Process (AIA) – A risk assessment process intended to support decision making regarding the responsible use of AI technologies. The AIA evaluates the areas of risk and any required mitigation measures to manage identified impacts. The results of the AIA will determine if the proposed AI technology is approved or not for the scope of its use (e.g., internal usage public facing services use, etc.).

The AIA aligns with the Government’s Guiding Principles in the Responsible Use of Artificial Intelligence by encompassing the following:

Transparent and Explainable: People are informed about when and how automation is used and are provided with meaningful information about decisions affecting them.
Fair and Ethical: Minimizes potential impacts on bias and discriminatory outcomes to ensure inclusion of all abilities.
Private and Secure: The collection, use, and disclosure of personal, sensitive, and confidential information for use with AI technologies as permitted under relevant privacy legislations.
The necessary controls are implemented to be secure and resilient to attack.
Accountable and Responsible: Stakeholders are accountable for AI technologies and governance mechanisms to ensure legal compliance.
Human Oversight: Humans remain in control of information created and decision making.

Generative AI (GenAI) – A type of AI that produces new content such as text, audio, code, videos, and images. GenAI technologies are typically trained on large volumes of data, including publicly available data. Based on the training data, GenAI can produce outputs that are statistically likely and appear to have been produced by humans.

Predictive AI – A type of AI that is focused on the future. It looks at existing data and uses it to make predictions and recommendations. You will see predictive AI in action for forecasting, optimization, and assessments.

Individual – For the purposes of OCIO AI policy instruments, the definition of individual refers to all employees, contractors, consultants, partners, students, volunteers, vendors, agents, third parties and other persons working on behalf of the Government of Newfoundland and Labrador.

The table below includes common abbreviations used by OCIO and acronyms found in this document.

Abbreviation	Description
AI	Artificial Intelligence
AIA	Artificial Intelligence Assessment Process
ATIPP	Access to Information Protection and Privacy
ATIPPA, 2015	Access to Information and Protection of Privacy Act, 2015
GenAI	Generative AI
MOIA	Management of Information Act
OCIO	Office of the Chief Information Officer
CIO	Chief Information Officer

^ Top of Page

4.0 Statements

Individuals shall not use AI technology unless it has been approved by Government and is used in compliance with this policy.
Individuals shall not enter government information, including privileged, confidential, and personal information, into publicly available AI technologies (e.g. ChatGPT).
Individuals can only input publicly available information into publicly available AI technologies.
Cabinet Records, as defined under MOIA, are not permitted to be captured, used, or managed in any AI technologies, except as may be explicitly authorized by Cabinet Secretariat.
Any information generated by approved AI technologies must be validated by an individual for accuracy before use and noted when used.
An individual must verify any proposed decision by a predictive AI technology before taking action to implement the decision.
Where the use of AI enabled technology involves the use, collection, or disclosure of personal information a Privacy Impact Assessment shall be completed by the individual/department as required under the ATIPPA, 2015.
This policy provides authority to the OCIO to establish additional policy instruments under this policy (directives, standards, and guidelines).
Exemptions related to this policy shall be approved by the CIO prior to AI use.

^ Top of Page

5.0 Roles and Responsibilities

Individuals are responsible for:

The accuracy of the information produced or the decision proposed by AI technology.

The Departments are responsible to:

Obtain Deputy Minister approval before submitting requests for use of AI technology to the OCIO.
Ensure all individuals working with the department are aware of this policy.
Monitor and evaluate information used and/or produced by AI technologies.
Develop processes to monitor and evaluate information used and/or produced by AI technologies.
Validate the data collected for, and used by, predictive AI to ensure it is relevant, accurate, up-to-date, and in accordance with applicable legislation, policies, directives etc. (e.g. MOIA, ATIPPA, 2015, etc.)

The OCIO is responsible to:

Develop, implement, and maintain this policy.
Make the approved AI Technologies available to all Government departments.
Collaborate with relevant stakeholders to review and evaluate requests for the use of AI technologies within Government departments based on the AIA.
Educate and increase awareness of this policy across government departments.
Provide training to Government departments on the responsible use of AI.
Maintain inventory of approved AI technologies including departmental use case.
Inform government departments when technology solutions use AI.
Obtain approval from CIO for all AI assessments completed under this policy and any issuance of exemptions related to this policy.
Collaborate with relevant stakeholders to review and evaluate requests for the use of AI technologies within Government departments based on the AIA.

The ATIPP Office is responsible to: