Somebody has to decide whether an AI tool is safe to use, be responsible for that decision and stand behind it; that’s a difficult call to make without the right knowledge or understanding of the stakes. This module covers core principles of AI security, including data protection, privacy, and governance, and provides the principles and checklist to make that call.
This is Module 2 of CAIAI’s Introduction to Artificial Intelligence course, available as an individual module covers the safety concerns that come with AI systems across different use cases, with a focus on data protection, privacy and governance. It also covers the technical and ethical side of hallucinations and deepfakes, and how to build strategies that identify and mitigate those risks.
The module runs across two chapters:
The specific ground includes AI security, cloud-based AI, local AI, AI hallucination mitigation and deepfake recognition. The cloud-versus-local question is also addressed, and this module examines that tradeoff.
The materials consist of chapter readings from the Introduction to Artificial Intelligence textbook, course videos, and online discussions where you post at least one comment addressing each chapter’s prompt. Both chapters close with a quiz requiring 70% or higher. Nothing is delivered live, so there is no session to miss, and you complete readings and videos at your discretion.
AI tools arrived before the governance did. Statistics Canada found that 19.2% of Canadian businesses used AI to produce goods or deliver services in the second quarter of 2026, triple the 6.1% recorded two years earlier, and among those businesses, only 32.0% provided AI-related training for existing employees. Most of the people putting company data into an AI tool were never told what happens to it.
The regulatory position is now explicit. Canada’s federal, provincial and territorial privacy regulators have published joint principles for responsible, trustworthy and privacy-protective generative AI technologies, setting out how existing privacy law applies to these tools.
In addition, concerns around hallucinated outputs are growing as organizations are held accountable for their AI outputs. This should be a major concern, as Statistics Canada reports that 31.4% of Canadian workers use generative AI daily, meaning many unverified AI-generated outputs are likely moving through Canadian workplaces every week.
Security is the part of our teaching where the questions get most specific to the room, as different organizations all have different data and different obligations attached to it. Below are just a few of those engagements that dove into their obligations as they pertain to AI:
These articles cover the same ground the module does, and give a fair preview of how it is taught.
Learners or employers are usually unaware of opportunities to pay for their learning. We have shared resources to help you identify available funding opportunities, including scholarships for individuals and grant programs for individuals and employers.
Apply for a scholarship through our Canadian AI scholarships page. In addition, visit our Canadian AI grant funding page to see what individuals or employers can apply for.
You can enrol directly on our learning platform, or use the form below if you want to ask any additional questions.
Our mission is to help Canadians learn, adapt and succeed with AI. Security is where succeeding and failing look identical right up until they do not. This module exists so that the person asked whether a tool is safe has something better than a guess to answer with.
