AI-900 retired on June 30, 2026. To earn Azure AI Fundamentals now, prepare for AI-901 using its current AI concepts and Microsoft Foundry implementation skills. This guide explains what changed, gives you an original Python exercise, and outlines study and exam-day preparation. Existing Fundamentals credentials do not expire under Microsoft’s current policy.
Confirm your target exam before you study
Exam AI-900 retired on June 30, 2026, and the Azure AI Fundamentals certification is now earned by passing Exam AI-901.
The old AI-900 study guide states that its exam retired on June 30, 2026, at 11:59 PM Central Standard Time. A remaining AI-900 practice-assessment listing does not establish that the exam can still be booked.
The continuing credential page names AI-901 as the required exam. Use its English skills list dated April 15, 2026. Localized versions generally follow English updates about eight weeks later, with variation; check the guidance for your booked language.
Official sources checked: Study guide for Exam AI-900: Microsoft Azure AI Fundamentals | Microsoft Learn — Retirement warning, June 30, 2026; Microsoft Certified: Azure AI Fundamentals - Certifications | Microsoft Learn — Certification details, required exams, registration notes; Exam AI-901: Microsoft Azure AI Fundamentals - Certifications | Microsoft Learn — Exam details, passing score, skills measured; Practice Assessments for Microsoft Certifications | Microsoft Learn — Overview and availability
What changed for AI-901?
AI-901 measures two skill areas: identify AI concepts and capabilities (40-45%) and implement AI solutions by using Microsoft Foundry (55-60%).
The current outline includes Foundry implementation tasks, so a preparation plan built only around the retired AI-900 service descriptions is incomplete. The percentages are official ranges, not exact question counts or required study-hour allocations.
Microsoft’s audience profile expects knowledge of Python syntax and programming techniques, familiarity with Azure resources, and conceptual knowledge of AI solutions in Azure. The exam page also identifies REST APIs, SDKs, and CLIs. Use that profile to assess gaps rather than assuming the old non-coding preparation advice still applies.
Official sources checked: Study guide for Exam AI-901: Microsoft Azure AI Fundamentals | Microsoft Learn — Skills measured as of April 15, 2026; Exam AI-901: Microsoft Azure AI Fundamentals - Certifications | Microsoft Learn — Exam details, passing score, skills measured; Study guide for Exam AI-900: Microsoft Azure AI Fundamentals | Microsoft Learn — Retirement warning, June 30, 2026; Microsoft Certified: Azure AI Fundamentals - Certifications | Microsoft Learn — Certification details, required exams, registration notes
AI concepts and capabilities: 40–45%
This area covers responsible AI considerations, AI model components and deployment choices, and the identification of common AI workloads.
Review the six responsible AI considerations: fairness, reliability and safety, privacy and security, inclusiveness, transparency, and accountability. As a study task, explain each and describe how several might apply to the same fictional application; a single cue word may not capture the whole requirement.
Review generative model components, model capabilities, deployment options, and configuration parameters. Practice explaining the information you would need before selecting a model or deployment.
The workload topics include generative and agentic AI; text extraction, entities, sentiment and summarization; speech recognition and synthesis; vision and image generation; and information extraction from text, images, audio, and video. Check the full current list when comparing older notes.
Official sources checked: Study guide for Exam AI-901: Microsoft Azure AI Fundamentals | Microsoft Learn — Skills measured as of April 15, 2026
Foundry implementation: 55–60%
The largest area tests building generative apps and agents, text and speech applications, vision-capable apps, and information extraction, all through Foundry.
The objectives include portal work and lightweight applications. Practice the listed tasks using current official learning material rather than memorizing SDK method names from an older course.
Review creating system and user prompts, deploying models, and interacting with models in the Foundry portal. Then work through the related client and agent tasks below.
- Generative apps and agents: build a chat client with the Foundry SDK, and create and test a single-agent solution plus a client app for an agent.
- Text and speech: build a text analysis app, respond to spoken prompts with a deployed multimodal model, and use Azure Speech in Foundry Tools.
- Vision: interpret visual input through a deployed multimodal model, generate new visuals, and build a lightweight app with vision capabilities.
- Information extraction: use Azure Content Understanding in Foundry Tools across documents and forms, images, audio, and video, and build a lightweight extraction app.
Official sources checked: Study guide for Exam AI-901: Microsoft Azure AI Fundamentals | Microsoft Learn — Skills measured as of April 15, 2026
Practice: filter labeled records with Python
Read this small list comprehension and predict its output before running it.
The audience profile names Python syntax and programming techniques. This original exercise practices tuple unpacking and filtering with fictional local data; it requires no cloud account or service call.
It is one syntax exercise, not a complete implementation study plan or an actual exam question. The categories are supplied inputs, so the code does not perform entity recognition or demonstrate model accuracy.
records = [("Ada", "person"), ("Cedar", "organization"), ("Lin", "person")]
names = [text for text, category in records if category == "person"]
print(names)Practice exercise
The code block supplies the complete records list. What exact output will print, which input record is excluded, and why?
Show answer
The output is ['Ada', 'Lin']. Cedar is excluded because its supplied category is organization, not person.
The comprehension visits the tuples in order, unpacks each into text and category, keeps only the tuples whose category equals "person", and collects their text values. Ada and Lin satisfy that condition. No labels are inferred, and no external data or model is involved.
Official sources checked: Study guide for Exam AI-901: Microsoft Azure AI Fundamentals | Microsoft Learn — Skills measured as of April 15, 2026; 5. Data Structures - Python 3.14.7 documentation — List comprehensions
Your preparation checklist
Use the AI-901 study guide as your map, take the official Practice Assessment, and explore the exam sandbox before test day.
Microsoft links the current practice assessment through AI Skills Navigator, which requires sign-in. Its questions are examples, not actual exam items, and do not represent the exam’s full length or complexity. Review the explanations and linked learning material instead of treating a result as a pass prediction.
- Compare your resources with the April 15, 2026 AI-901 skills list, especially the Foundry implementation tasks.
- Identify gaps in Python, Azure resources, REST APIs, SDKs, and CLIs using the audience profile.
- Work the local Python example, then use official learning material for the listed Foundry app and agent tasks.
- Use the current practice assessment to find topics needing further study.
- Explore the sandbox and review break and navigation rules.
- Request accommodations in advance if needed and check booking instructions.
Official sources checked: Microsoft Certified: Azure AI Fundamentals - Certifications | Microsoft Learn — Certification details, required exams, registration notes; Study guide for Exam AI-901: Microsoft Azure AI Fundamentals | Microsoft Learn — Skills measured as of April 15, 2026; Exam AI-901: Microsoft Azure AI Fundamentals - Certifications | Microsoft Learn — Exam details, passing score, skills measured; Practice Assessments for Microsoft Certifications | Microsoft Learn — Overview and availability; Exam duration and exam experience | Microsoft Learn — Fundamentals exam duration table, breaks, Learn access
What to expect on exam day
Fundamentals exams give you 45 minutes of exam time inside a 65-minute seat booking, and a scaled score of 700 or greater passes.
The 65-minute seat time covers instructions, the candidate agreement, the 45-minute assessment, and comments. The standard exam duration already includes the additional five minutes provided for possible breaks; do not add five more.
Microsoft’s general guidance says most certification exams have 40–60 questions, with variation. This does not establish a fixed AI-901 count or item mix. Use the sandbox to learn the interface.
The passing score is 700 on a scale of 1–1,000, not a raw 70% requirement. Incorrect answers do not deduct points. Some questions may be unscored and are not identified, so answer every question.
The exam is proctored, and Microsoft Learn is unavailable during Fundamentals exams. A break does not pause the clock, and you cannot return to previously viewed questions, including flagged or unanswered ones. Review those questions before starting a break.
Official sources checked: Exam AI-901: Microsoft Azure AI Fundamentals - Certifications | Microsoft Learn — Exam details, passing score, skills measured; Exam duration and exam experience | Microsoft Learn — Fundamentals exam duration table, breaks, Learn access; Exam scoring and score reports | Microsoft Learn — Scores needed to pass exams
Booking and retakes
Check the fee for your testing region and use the current AI-901 scheduling link.
Pricing depends on the country or region of proctoring. Microsoft recommends registering with a personal account so records remain accessible if you leave a school or employer.
After a first failure, wait 24 hours before the second attempt. A 14-day wait applies between later attempts, with at most five attempts in 12 months from the first. Retake fees apply where required; the waiting period does not guarantee appointment availability.
If the exam is unavailable in your preferred language, Microsoft describes a process to request an additional 30 minutes. Check localized update guidance and request any accommodations in advance.
Official sources checked: Exam AI-901: Microsoft Azure AI Fundamentals - Certifications | Microsoft Learn — Exam details, passing score, skills measured; Microsoft Certified: Azure AI Fundamentals - Certifications | Microsoft Learn — Certification details, required exams, registration notes; Exam and Applied Skills assessment lab retake policies | Microsoft Learn — Retake waiting periods; Study guide for Exam AI-901: Microsoft Azure AI Fundamentals | Microsoft Learn — Skills measured as of April 15, 2026
Official sources checked
Verified exam facts:
- Microsoft Certified: Azure AI Fundamentals - Certifications | Microsoft Learn
- Study guide for Exam AI-901: Microsoft Azure AI Fundamentals | Microsoft Learn
- Study guide for Exam AI-900: Microsoft Azure AI Fundamentals | Microsoft Learn
- Exam AI-901: Microsoft Azure AI Fundamentals - Certifications | Microsoft Learn
- 5. Data Structures - Python 3.14.7 documentation
- Exam duration and exam experience | Microsoft Learn
- Exam scoring and score reports | Microsoft Learn
- Credential expiration policies | Microsoft Learn
- Exam and Applied Skills assessment lab retake policies | Microsoft Learn
- Practice Assessments for Microsoft Certifications | Microsoft Learn