If you have been putting off studying for Microsoft’s Azure AI Fundamentals certification because you were not sure where to start, now is a good time to jump in. The AI-900 exam has officially been retired, and it has been replaced by the AI-901, and the changes are bigger than a name swap. I just published the first video in a new training series that walks through exactly what is different, and this post gives you the highlights before you press play.

Why the AI-901 Matters Right Now

AI is no longer a side topic in IT, it is becoming a core part of nearly every role, and that is especially true for anyone working in IT. The AI-901 is Microsoft’s entry-level certification for Microsoft AI, designed to build a solid foundation before you move on to more advanced certifications. Whether you are brand new to AI or have some hands-on experience, this exam is meant to close knowledge gaps and give you a baseline you can build on. It also happens to be one of the few fundamentals-level exams that includes real, hands-on experience with a Microsoft platform, in this case, Microsoft Foundry.

What Changed from AI-900 to AI-901

The AI-900 was built around five knowledge domains, and all of it was theory-based. There was no hands-on requirement anywhere in the exam. The AI-901 simplifies that structure into two domains, but do not let the smaller number fool you, the scope of the exam actually grew.

The first domain, Identify AI Concepts and Capabilities, makes up 40 to 45 percent of the exam and remains largely theory-based. Most of the foundational concepts from AI-900 still show up here, just reorganized. One notable change is that the dedicated Machine Learning Fundamentals domain from AI-900 has been removed entirely. In its place, generative AI and large language model knowledge are now treated as baseline skills every Azure AI professional needs.

The second domain, Implement AI Solutions by Using Microsoft Foundry, accounts for 55 to 60 percent of the exam, and this is where things get interesting. This section is skill-based, and it requires real experience working inside Microsoft Foundry, not just reading about it.

Why Microsoft Foundry Sits at the Center of the Exam

Microsoft Foundry is now the central hub for Azure AI development. It is a single platform where you can build, test, deploy, and manage AI solutions, work with AI models, and create intelligent agents, all from one place. Instead of bouncing between disconnected AI services, Foundry brings everything together, including access to thousands of models from Microsoft, OpenAI, Anthropic, Meta, and other providers.

Microsoft has made Foundry the primary platform for building AI solutions, and the AI-901 reflects that decision directly. The exam now tests your ability to actually work inside Foundry, not just describe what it does. That shift is part of a bigger strategy of consolidating AI development tools into a single platform, and validating practical skills instead of testing concepts alone.

Agentic AI: The Exam’s Newest Focus

If your experience with AI so far has been limited to chatbots that take a prompt and return a response, agentic AI is a meaningful step beyond that. Agentic AI systems can plan, reason, make decisions, and carry out multi-step tasks across multiple tools with minimal human intervention. Individual agents can also be connected to specific knowledge sources.

A simple way to picture this is a corporate chatbot. If one chatbot tried to answer every question using every document across an entire organization, it would be slow, and results would often be irrelevant or inaccurate. With agentic AI, you can build a dedicated IT support agent connected only to IT documentation and tickets, and a separate HR agent connected only to HR policies. When someone asks an IT question, the system routes it to the IT agent. When someone asks about employment policy, the HR agent takes over. That narrow scope improves speed, relevance, and accuracy, and it is a great example of agentic AI doing real work.

Generative AI and Large Language Models, the Building Blocks

Behind the scenes, agentic AI relies heavily on generative AI and large language models, or LLMs, to reason, generate responses, and interact with users. Generative AI creates new content rather than simply analyzing existing information, which is what separates it from traditional machine learning models. A large language model is a specific type of generative AI model that focuses on language, trained on massive amounts of text so it can write, summarize, answer questions, and hold a conversation.

Understanding how LLMs actually generate that language comes down to a few key exam terms: tokens, training, and inference. Tokens are the small chunks of text a model breaks language into as it predicts what comes next. Training is the one-time, resource-intensive process where a model learns patterns from enormous datasets. Inference is what happens every time you actually use the model, sending a prompt and getting a generated response back. Training is how the model learns, inference is how it responds, and both concepts show up directly on the AI-901.

IMAGE TOKENS

What’s Coming Next in This Series

This video is the foundation for the entire AI-901 training series. Future videos will dig into Domain 1 topics like responsible AI, choosing the right AI model, and how AI understands text, speech, and vision. From there, the series moves into Domain 2, with hands-on Microsoft Foundry builds including chat applications, AI agents, computer vision, and Azure Content Understanding, eventually combining everything into one complete AI-powered solution.

Watch the Full Video

Reading about the AI-901 is a good start, but watching the video is where these concepts really click, especially the side-by-side breakdown of AI-900 versus AI-901 and the walkthrough of how Microsoft Foundry fits into everything. Watch the full video now, and subscribe so you do not miss the rest of the series as it goes deeper into Azure AI Fundamentals. Don’t forget to check out the resources in the links below for more information.

Links:

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 Certifications poster
https://learn.microsoft.com/en-us/credentials/