Public Service Policy 4-16
Acceptable AI Use

Shelves of books at the Eastside Branch

Acceptable AI Use Policy

 

Purpose

Waterloo Public Library recognizes the opportunities and responsibilities inherent in using emergent technology. This policy outlines the acceptable use and application of AI tools approved by Waterloo Public Library.

Scope

This policy provides a framework for the responsible, human-centered use of AI by all staff, volunteers and consultants across all departments. It covers both internal use of AI and the use of AI in the delivery of public services.

Definitions

“Agentic AI” are artificial intelligence systems that can accomplish goals with limited human supervision. Agents mimic the processes human beings use to make decisions in real time.

“Artificial Intelligence (AI)” is the ability of a computer or computer-controlled robot to perform tasks commonly associated with intelligent beings. AI can include both computer software and physical machines and performs tasks commonly associated with human beings such as learning, reasoning, problem solving, perception and language.

“Generative AI” is a subset of AI that creates new content in varied forms (text, images, audio, code etc.). Gen AI models learn underlying patterns and structures from the data they are trained on and use that information to produce new data outputs often in response, to natural language prompts.

“Personal Information Data” is about an identifiable individual. Personal data includes identifiable information about staff, clients, customers, volunteers, donors, and partners.

“Sensitive Data” includes personal data related to financial, legal or other sensitive areas identified in Waterloo Public Library policies. This includes donation records, payroll and accounting information, intellectual property, board communications, contracts and privileged correspondence.

Policy Statement

Waterloo Public Library holds five core organizational values: Trust, Equity, Joy of Discovery, Future Readiness and Compassion. These values guide our organizational use of artificial intelligence.

Trust

As a pillar of information literacy and an organization that is trusted by its community, Waterloo Public Library approaches the use of AI with a decisive ‘human at the helm’ lens. No AI tool will be used where there is not human initiation to begin an activity and human scrutiny over what is produced.

The Library will leverage the opportunities AI presents to improve and enhance workflows without compromising our professional standards of ethical, transparent and accountable information use.

Waterloo Public Library and its staff will be prepared to disclose when and how AI is used.

Equity and Compassion

The Library will actively work to create equitable access to opportunities for the community to learn about and use AI.

AI will not be used in ways where its biases or outputs are likely to result in real harm to humans. This includes use of AI tools for final decision-making in contexts like hiring, human resource matters or service planning.

Future Readiness and Joy of Discovery

Staff will have intentional opportunities to learn about, explore, and apply AI in their work.

The Library will explore opportunities to automate work that can be done by AI so staff themselves can more fully prioritize human-centered service delivery.

Staff Responsibilities

All staff, volunteers and contractors using AI on behalf of Waterloo Public Library must:

  • Complete relevant AI training before using AI in their work.
  • Use AI tools for appropriate purposes within their role.
  • Comply with data privacy requirements of the Municipal Freedom of Information and Protection of Privacy Act, R.S.O. 1990 (MFIPPA) and all WPL policies.
  • Disclose use of AI to members of the public when AI is used to provide a service to them OR makes use of information connected to them.
  • Treat AI as a tool, not an expert. AI outputs generated by anyone acting on behalf of Waterloo Public Library must be reviewed, assessed and validated by a human being before being used. Staff remain accountable for the content of their work when it is generated by AI.
  • Safeguard intellectual property and copyrighted material. This includes avoiding inputting the intellectual property of anyone but the user into an AI tool. This also includes purchasing or recommending AI produced resources without disclosing to the potential consumer that the item was produced by AI.
  • Refrain from using AI in ways that have the potential to cause harm through bias, discrimination, disclosure of private information, threaten safety or provide wrong, misleading or poorly produced information.

Data Governance

All staff, volunteers and contractors using AI on behalf of Waterloo Public Library will:

  • Follow organizational data protection protocols when inputting data into AI tools.
  • Avoid entering Personal Information Data (PID) or Sensitive Data into AI tools that have not been approved for such purpose.
  • Monitor the storage of AI outputs (chat logs, datasets etc.) applying record maintenance protocols to this data in accordance with WPL record management and privacy policies.
  • Conduct privacy impact assessments in cases where approved enterprise tools that process personal information integrate AI (ex. Library catalogue system).

Acceptable Uses

Within the responsibilities outlined above, staff, volunteers and contractors of Waterloo Public Library may use AI in the following ways.

Creative Brainstorming activities like generating new ideas for programs and services, listing potential steps to consider when outlining a project charter, creating content outlines, and generating potential interview questions.

Community Education activities like programming for all ages to build AI digital literacy, customer support using AI enhancements, facilitating public access to AI tools.

Content Revision and Editing activities like reviewing and editing communications like program descriptions, grant narratives, marketing materials and reports.

Customer Experience Enhancements activities like recommendation models for collection and program materials of interest, accessibility enhancements.

Language Translation for non-sensitive materials such as library and community resources, equipment and collection information.

Productivity Enhancements to workflows or tools authorized by the corporation. Examples include email management and meeting transcription.

Research and Information Gathering including article summaries, compiling information, gathering samples of comparable services, policies, programs, etc. suggesting insights and trends within anonymized data sets from surveys where PID has been removed.

Technical Support including assisting with coding and automation of internal processes, cyber threat and fraud detection tools.

Training and Onboarding including formatting training resources, creating instructional videos, generating application activities, etc.

Other applications may be considered after appropriate review and approval by a supervisor or division manager.

Prohibited Uses

Waterloo Public Library will not make legal, hiring, financial or other high risk decisions using AI tools.

Outside of secured and approved tools, staff will not knowingly enter Personal Information Data about staff or customers into AI tools. Anyone acting on behalf of Waterloo Public Library is prohibited from entering sensitive or confidential information obtained through the course of doing the library’s work into any open AI system.

Staff will not respond to reference questions or information requests using AI without checking the results against vetted sources.

No staff member will conduct surveillance or profile individuals using AI in any way without disclosing this activity.

Staff will not use AI tools without first critically evaluating the tool according to established procedures for doing so.

Document Record

September 2025 – Initial Release