epifive

Posted on Oct 26 Originally published at epifive.com

AI-900 Retired: AI-901 Azure AI Fundamentals Study Guide and Plan

#ai900 #ai901 #azureai #microsoftlearn

Originally published on EpiFive, where it is kept up to date.

AI-900 can no longer be taken: Microsoft retired it on 30 June 2026, and the Microsoft Certified: Azure AI Fundamentals certification is now earned by passing exam AI-901. AI-901 keeps the fundamentals level but moves from describing Azure AI services to building with them, so about 55 to 60 percent of the exam is implementing solutions in Microsoft Foundry, and Python syntax and SDK use are now part of the expected background.

This guide is for people who searched for an AI-900 study plan and need to know what to study instead, and for anyone starting Azure AI Fundamentals. By the end you will know what changed, what the AI-901 skills outline covers, which free resources Microsoft points to, and how to organise two to three weeks of practice. Details come from Microsoft's AI-900 and AI-901 study guides and the certification and exam pages on Microsoft Learn, read on 25 October 2026. Exams change, so check the current study guide before you book.

What happened to AI-900

Microsoft's AI-900 study guide opens with a warning that the exam was retired on 30 June 2026. The certification page now lists AI-901 as the required exam, with the job role AI Engineer. Both the AI-900 study guide and the retired outline are still published, so older articles and videos may describe skills that are no longer assessed.

Item	AI-900 (retired)	AI-901
Status	Retired 30 June 2026	Current; no retirement date listed
Passing score	700	700
Skills outline version	As of 2 May 2025	As of 15 April 2026
Main emphasis	Describing AI workloads, machine learning, vision, NLP and generative AI on Azure	AI concepts, plus implementing AI apps and agents with Microsoft Foundry
Background expected	Basic cloud concepts, client-server applications; no data science or coding experience needed	Conceptual knowledge of Azure AI, Python syntax and programming techniques, familiarity with Azure resources

If you already studied for AI-900, much of the conceptual material still helps, but plan time for the hands-on side.

AI-901 at a glance

The exam has two skill areas.

Skill area	Weight
Identify AI concepts and capabilities	40-45%
Implement AI solutions by using Microsoft Foundry	55-60%

The pass mark is 700. The exam is offered in English, Arabic (Saudi Arabia), Chinese (Simplified and Traditional), French, German, Indonesian, Italian, Japanese, Korean, Portuguese (Brazil), Russian and Spanish. Price depends on the country or region where it is proctored. Microsoft recommends registering with a personal Microsoft account, because exam records tied to a work or school account are lost if you leave the organisation.

Microsoft also notes that most questions cover generally available features, preview features may appear if commonly used, and you should be familiar with REST APIs, SDKs and CLIs.

Skills measured in detail

Identify AI concepts and capabilities (40-45%)

Describe principles of responsible AI:

Fairness, reliability and safety, privacy and security, inclusiveness, transparency and accountability

Identify AI model components and configurations:

How generative AI models work
Choosing an appropriate model based on capabilities
Model deployment options and configuration parameters

Identify AI workloads:

Scenarios for common workloads, including generative and agentic AI, text analysis, speech, computer vision and information extraction
Text analysis techniques: keyword extraction, entity detection, sentiment analysis and summarisation
Speech recognition and speech synthesis
Computer vision and image-generation models
Extracting information from text, images, audio and video

Implement AI solutions by using Microsoft Foundry (55-60%)

Generative AI apps and agents:

Write effective system and user prompts
Deploy a model and interact with it in the Foundry portal
Create a lightweight chat client application with the Foundry SDK
Create and test a single-agent solution in the Foundry portal
Create a lightweight client application for an agent

Text and speech:

Build a lightweight application that includes text analysis
Respond to spoken prompts using a deployed multimodal model
Build a lightweight application using Azure Speech in Foundry Tools

Computer vision and image generation:

Interpret visual input in prompts using a deployed multimodal model
Create new visual outputs with generative models
Build a lightweight application with vision capabilities

Information extraction:

Extract information from documents and forms using Azure Content Understanding in Foundry Tools
Extract information from images, and from audio and video, with Content Understanding
Build a lightweight application with information extraction capabilities

Note what is no longer listed. The AI-900 outline had a dedicated machine learning domain with regression, classification and clustering and an Azure Machine Learning section. AI-901's published outline does not list those as separate skills, so treat them as background, not as exam targets, and follow the current outline.

Free and official resources

The AI-901 exam page: it links the study guide and the exam sandbox. The certification page said learning paths and modules were not yet available when checked, so look there for new official training.
The AI-901 study guide: the authoritative list of skills and their weights.
The exam sandbox, to try the question types and the interface.
The practice assessment on AI Skills Navigator, on the Azure AI Fundamentals certification page. You must be signed in to launch it.
Documentation the study guide lists, such as Azure Machine Learning, Computer Vision, Speech and Azure Bot Service. Some of these links point to older product names, so confirm the service names you meet in Foundry.

For hands-on practice with a deployed model, the existing articles on Azure OpenAI deployment types and fixing 429 rate limit errors show what deploying and calling models looks like in practice. They are not exam material, but they build familiarity with deployments, parameters and limits.

A 3-week practice plan

The plan assumes about an hour a day, and some Python comfort. If you are new to Python, add a week for basics.

Week	Focus	Practice
1	AI concepts and capabilities: responsible AI principles, how generative models work, choosing and configuring models, the main workloads	Write a one-line definition of each responsible AI principle and one example scenario for each workload
2	Foundry basics: deploy a model, prompt it in the portal, build a lightweight chat client with the SDK, create and test a single agent	Use a trial or Azure free account; deploy a model, try different system prompts and parameters, then call it from a small script
3	Text, speech, vision and information extraction in Foundry, then review	Build one small app for each: text analysis, speech, an image prompt, and document extraction; take the practice assessment and review gaps

Study habits that help

Practise tasks rather than only reading. More than half the exam is about implementing, so click through the portal and run code.
Keep a table of workload, typical input, typical output and the Foundry capability that handles it.
Learn the responsible AI principles well enough to recognise a scenario that violates each one.
After the practice assessment, study the objectives you missed instead of retaking it until questions repeat.
Watch for outdated materials. Anything that teaches AI-900 domains such as clustering or Azure Machine Learning designer in depth may no longer match the exam.

Before exam day

Read the exam policies and FAQs, check your ID and proctoring requirements, and try the sandbox. If you do not pass, the Microsoft retake policy applies; read it on Learn before rebooking.

Checklist

Confirm you are studying AI-901, not the retired AI-900
Read the current AI-901 study guide and note the 15 April 2026 version
Cover responsible AI, model selection and AI workloads
Deploy a model, build a chat client and test a single agent in Foundry
Build small apps for text, speech, vision and information extraction
Take the practice assessment on AI Skills Navigator
Register with a personal Microsoft account

References

This guide was researched and drafted with AI assistance and checked against the vendor documentation listed in its References. Read the maintained version and related guides on EpiFive.

DEV Community