If an AI chatbot makes a promise your company can’t keep, who’s responsible? In 2024, a Canadian tribunal answered that question when Air Canada argued its chatbot was responsible for its own words. The tribunal disagreed, and that case gets to the heart of one of Microsoft’s six principles of responsible AI. The latest video in my AI-901 Azure AI Fundamentals series breaks down all six, and this post gives you a preview before you press play.

Why Responsible AI Is on a Technical Exam

Responsible AI can sound like something only the legal team worries about, but many of the decisions that put it into practice are technical ones. Which content filters are enabled, what data an agent can access, what gets logged, and who reviews the output all shape how responsible an AI solution really is. That’s why responsible AI is part of the AI-901 exam, and why the exam expects you to look at a scenario and recognize which principle applies.

Six Principles, Real-World Examples

The video walks through fairness, reliability and safety, privacy and security, inclusiveness, transparency, and accountability, and pairs each one with a real example of what happens when it’s missing. You’ll see how an experimental recruiting tool learned to downgrade resumes, why Microsoft shut down its own Tay chatbot in 2016, how a confidential code leak led one company to ban generative AI tools, and why “the AI did it” isn’t a defense.

Along the way, you’ll also learn why generative AI produces hallucinations, and how grounding and retrieval augmented generation (RAG) help keep AI responses tied to trusted data.

Responsible AI in Microsoft Foundry

The video goes beyond the definitions to show how Microsoft puts these principles into practice. You’ll get a look at guardrails in Microsoft Foundry, including content safety, Prompt Shields for jailbreak and prompt injection attacks, and built-in evaluators. You’ll also see how Microsoft’s approach to managing AI risk aligns with the NIST AI Risk Management Framework, and why governance matters more as generative and agentic AI systems grow.

Watch the Full Video

Reading about responsible AI is a good start, but the video is where these principles come together, with real-world examples, a look inside Microsoft Foundry, and exam tips for matching any scenario to the right principle. Watch the full video now, and subscribe on YouTube so you don’t miss the next video in the AI-901 series, where we look at how to pick the right AI model and identify AI workloads.

Links:

A Beginner’s Guide to the AZ-900
https://www.udemy.com/course/beginners-guide-az-900/?referralCode=C74C266B74E837F86969

Zero to Hero with Azure Virtual Desktop
https://www.udemy.com/course/zero-to-hero-with-windows-virtual-desktop/?referralCode=B2FE49E6FCEE7A7EA8D4

Hybrid Identity with Windows AD and Azure AD
https://www.udemy.com/course/hybrid-identity-and-azure-active-directory/?referralCode=7F62C4C6FD05C73ACCC3

Windows 365 Enterprise and Intune Management
https://www.udemy.com/course/windows-365-enterprise-and-intune-management/?referralCode=4A1ED105341D0AA20D2E

Exam AI-901: Microsoft Azure AI Fundamentals
https://learn.microsoft.com/en-us/credentials/certifications/exams/ai-901/

Study guide for Exam AI-901: Microsoft Azure AI Fundamentals
https://learn.microsoft.com/en-us/credentials/certifications/resources/study-guides/ai-901

Microsoft Responsible AI Principles and Approach
https://www.microsoft.com/en-us/ai/principles-and-approach

Guardrails and controls overview in Microsoft Foundry
https://learn.microsoft.com/en-us/azure/foundry/guardrails/guardrails-overview

NIST AI Risk Management Framework
https://www.nist.gov/itl/ai-risk-management-framework