Available nowTake the course today

How Amii Uses AI

We believe AI can advance research and industry, so we are open about how our team responsibly uses these tools in our daily work.

Our principles

AI supports our people’s judgment. It doesn’t replace it.

AI is a powerful but imperfect tool. It can produce convincing errors, so it supports our people’s judgment rather than replacing it. Every Amii team member follows these principles:

Suggest, not solve

AI can offer ideas and options; people decide.

Summarize, not substitute

AI can condense material; people check it against the original source.

First draft, not final word

AI can help start a draft; people write, check and approve what we publish.

Edit, not editorialize

AI can polish wording; it does not set our positions or voice.

Direct, not abdicate

When AI tools take actions, staff give them a specific goal and clear limits, never open-ended control.

Verify, not assume

Staff check AI output for accuracy and review what an AI tool plans to do before it acts.

Question, not inherit

AI can reflect bias in the data it learned from; we check its output for unfair assumptions rather than passing them along.

Scope, not generalize

AI tools get only the access and information a task needs.

Supervise, not abandon

A person stays involved throughout and must approve any action that affects clients or the public, or that cannot easily be undone.

Disclose, not disguise

We are open about when and how AI was involved in our work.

Own, not offload

People stay accountable for every outcome; “the AI did it” is never an excuse.

Transparency

When we tell you AI was involved

Talking to AI

On our websites, you are talking to a person when you use live chat or email us. If we introduce an AI assistant on any Amii site, it will say clearly that it is AI at the start of every conversation, and you will always be able to reach a person via email instead.

Images, audio and video

We label images, audio and video that are substantially created by AI so they aren’t mistaken for real photographs or recordings.

Written content

Our staff use AI to brainstorm, research, summarize or edit. We do not label that text, because a person wrote, checked and approved it and stands behind it. If a piece is purposely AI-generated, for example, a demonstration in one of our courses, we say so.

Course examples

Our asynchronous courses often include examples of AI outputs, some of which contain errors on purpose so you can learn to spot them. These are always identified as AI outputs.

Protecting your information

We do not put personal information, confidential client information, contracts, financial records or recruitment materials into publicly accessible free AI tools. Any AI tool used with this kind of information must first pass our internal AI tool procurement review, led by our IT team.

Where AI services offer the choice, we opt out of having our data used to train their models. Before we connect an AI tool to our business systems, our IT and engineering leaders review it for security.

Our Privacy Policy explains how we collect and use personal information, and our Terms of Service explain how others may use the content on our websites, including for training AI models.

AI in client projects and research

Our work with industry follows Amii’s Principled AI Framework. When a project uses a commercial large language model or another third-party AI service, such as tokens, we discuss it with the client first and record their agreement.

For each project, a model card documents which models were used, what they are intended for, and their known limitations. When we collect data for a project, a datasheet documents its source and what it should and should not be used for. Client information is used only for that client’s project, and only in tools approved for that purpose.

Both official languages

Amii serves Canadians in English and French. We make our web pages available in both languages using professional automated translation software, selected for quality and able to handle scientific terms. Because automated translation can miss nuance, certified translators regularly proofread our content. We aim for continual improvement with biannual reviews of priority content.

If something in either language is unclear or wrong, tell us, and we will correct it.

Using AI thoughtfully

We use AI where it adds value. AI uses energy and computing resources. We use it where it genuinely improves our work, not by default, and we choose appropriately sized tools for the task.

When AI gets it wrong: Despite human review, mistakes can happen. If you believe something on an Amii website contains an AI-related error, or that we used AI in a way that does not match this page, please tell us (see Contact us below).

We will acknowledge your report within 15 business days, investigate, and correct any errors we confirm. We log AI-related errors and incidents internally, review their causes, and update our practices so they don’t happen again.

Standards and oversight

Our practices align with the Government of Canada’s Voluntary Code of Conduct on the Responsible Development and Management of Advanced Generative AI Systems, and are informed by the international ISO/IEC 42001 AI management standard and the NIST AI Risk Management Framework.

Our Board of Directors and Executive Leadership Team are accountable for how Amii uses AI, including affiliated policies and procedures. Any AI capability that could share information outside Amii requires executive authorization and approval each time it is used. We review our AI practices and this page at least once a year, and sooner when the technology or the law changes.

Contact us

Questions about how we use AI, or want to report an error? Email hello@amii.ca with “AI use” in the subject line, or write to Amii, #1101, 10065 Jasper Avenue, Edmonton, Alberta T5J 3B1. You can contact us in English or French.

Last updated: October 2026