The Short Answer
If you are not yet certified, take AI-901. Microsoft retired Exam AI-900 on June 30, 2026, and Exam AI-901 is now the only exam that earns the Microsoft Certified: Azure AI Fundamentals certification. You cannot schedule AI-900 anymore.
If you already hold the Azure AI Fundamentals certification from passing AI-900, you do not need to retake AI-901. Microsoft fundamentals certifications do not expire, and the certification itself was not retired — only the exam behind it changed.
The real question for most readers in 2026 is not really "which exam should I sit?" (you sit AI-901) but "what changed, does my old AI-900 knowledge still count, and do I need to do anything?" This comparison answers all three, with the official Microsoft Learn pages as the source of truth.
The Retirement Timeline
Microsoft's announcement and the official exam pages establish the timeline:
| Date | Event |
|---|---|
| April 15, 2026 | English version of AI-901 updated and published on Microsoft Learn |
| April 2026 | AI-901 beta opened to candidates |
| June 30, 2026 | AI-900 retired at 11:59 PM CST; no longer schedulable |
| June 2026 | AI-901 reached general availability |
| July 2026 (today) | AI-901 is the only current Azure AI Fundamentals exam |
The AI-900 exam page now carries a warning: "The certification requirements have changed. The AI-900 exam was retired on June 30, 2026, and has been replaced by AI-901. To earn this certification, candidates must now pass AI-901." The AI-901 exam page lists no retirement date.
A beta period preceded general availability. The beta is now closed; if you are scheduling today, you register for the generally available exam at standard regional pricing. Beta candidates received the same credential after results were scored.
What Actually Changed: AI-900 vs AI-901
The certification name is identical. The exam behind it is substantially different.
| Area | AI-900 (retired) | AI-901 (current) |
|---|---|---|
| Audience | Technical and non-technical beginners | Technical beginners planning to build AI solutions |
| Focus | "What is AI?" — describe and recognize | "How do I build an AI app using Microsoft Foundry?" |
| Domains | 5 | 2 |
| Required coding | None | Basic Python syntax and programming concepts |
| Largest domain | Generative AI workloads (20–25%) | Foundry implementation (55–60%) |
| Languages | 13 | English (others follow ~8 weeks after English updates) |
| Platform emphasis | Individual Azure AI services | Microsoft Foundry (unified platform) |
AI-900 tested whether you could describe AI concepts and identify when to use Azure AI services. AI-901 tests whether you can implement AI solutions inside Microsoft Foundry — the unified Azure platform for building, deploying, and managing AI apps and agents. The verb shift from "describe" to "implement" is the single biggest change, and it is why AI-901 requires basic Python while AI-900 did not.
The Content Domain Diff
This is where old AI-900 study plans diverge most sharply from AI-901. The official AI-901 study guide lists the current skills measured.
AI-900's five domains (each 15–25%):
- Describe AI workloads and considerations (15–20%)
- Describe fundamental principles of machine learning on Azure (15–20%)
- Describe features of computer vision workloads on Azure (15–20%)
- Describe features of NLP workloads on Azure (15–20%)
- Describe features of generative AI workloads on Azure (20–25%)
AI-901's two domains:
- Identify AI concepts and capabilities (40–45%)
- Implement AI solutions by using Microsoft Foundry (55–60%)
What is de-emphasized in AI-901: traditional ML subtopics as named exam areas (regression, classification, clustering), Azure Machine Learning Studio pipelines and automated ML, the Azure Bot Service framing (replaced by agentic AI), and individual Azure Cognitive Services naming (collapsed into Foundry services).
What is new in AI-901: AI agents and multi-step agentic workflows, Foundry as the unified hub, multimodal models including image generation, Azure Content Understanding for information extraction, prompt engineering as a tested skill, and the Foundry SDK.
The weighting tells you where to spend your prep time. Because Domain 2 is 55–60% and Domain 1 is 40–45%, roughly six of every ten questions touch Foundry implementation. If you study only concepts and skip hands-on Foundry work, you are leaving more than half the exam score on the table.
Common Misconceptions About the Retirement
Competing pages — many written before the retirement — leave readers with wrong impressions. Here are the four most common ones:
1. "The Azure AI Fundamentals certification was retired." Wrong. Only the AI-900 exam was retired. The certification is current and is now earned by passing AI-901. The credential on your transcript is the same whether you passed AI-900 in 2024 or AI-901 in 2026.
2. "AI-900 cert holders lose their certification." Wrong. Microsoft fundamentals certifications do not expire. Associate, expert, and specialty certifications require annual renewal; fundamentals do not. Your AI-900-earned credential stays valid permanently with no action required.
3. "AI-901 is just a renumbering of AI-900." Wrong. AI-900 was a describe-and-recognize exam with five domains and no coding. AI-901 is an implement-and-build exam with two domains, a 55–60% Foundry implementation section, and a basic Python prerequisite. The content, structure, and required skills are materially different.
4. "Old AI-900 study materials are now useless." Partly wrong. Domain 1 of AI-901 (concepts, responsible AI, workloads) overlaps with AI-900's conceptual content. Domain 2 (Foundry implementation) is largely new and needs fresh materials. See the next section for what transfers.
Do AI-900 Cert Holders Need to Retake?
No. Microsoft fundamentals certifications do not expire. The certification itself was not retired — only the exam required to earn it changed. Your existing Azure AI Fundamentals credential stays on your transcript, and you need take no action.
You may optionally take AI-901 if you want to showcase the broader Foundry and agentic AI skills to employers, but there is no requirement to do so and no recertification obligation. Treat AI-901 as a career-growth choice, not a compliance task.
Do AI-900 Study Materials Still Help?
Partially. This matters if you bought AI-900 books, courses, or practice sets and wonder whether they are now wasted.
- Domain 1 overlap is real. AI-901's "Identify AI concepts and capabilities" domain (40–45%) covers responsible AI principles, generative AI model basics, model selection, and the workload families (generative/agentic, text, speech, vision, information extraction). AI-900's conceptual coverage of responsible AI, vision, NLP, and generative AI transfers here — though the framing is narrower and updated.
- Domain 2 is largely new. The Foundry implementation domain (55–60%) has no real equivalent in AI-900. If your old materials focus on describing Azure Cognitive Services individually, they will not prepare you for Foundry portal, SDK, agents, or Content Understanding questions.
- Python is new. Any AI-900 prep that skipped coding entirely is insufficient for AI-901.
Use AI-900 materials for Domain 1 vocabulary and responsible AI principles, then invest fresh time in Microsoft's two official AI-901 learning paths and hands-on Foundry work for Domain 2. For the legacy exam content, see our Azure AI Fundamentals AI-900 exam guide.
Eligibility and the Python Gate
The AI-901 exam page describes the candidate as "at the beginning of your career in AI solution development" and lists:
- Conceptual knowledge of AI solutions in Azure
- Foundational technical skills to work with Azure AI solutions
- Knowledge of Python coding syntax and programming techniques
- Familiarity with Azure resources
AI-900 listed no coding prerequisite and explicitly welcomed non-technical candidates — its page stated the exam was "intended for you if you have both technical and non-technical backgrounds" and that "data science and software engineering experience are not required." AI-901's Python requirement is the most consequential eligibility change. You do not need to be a software engineer, but you should be comfortable reading short Python snippets — syntax, variables, function calls, and basic control flow. If you have never written Python, plan a few days of Python basics before starting the Foundry implementation modules.
Cost, Scheduling, and Format
Verified from the official AI-901 exam page:
- Passing score: 700 on a 1,000-point scale (same as AI-900).
- Language: English (updated April 15, 2026; localized versions follow approximately eight weeks later).
- Retirement date: None.
- Pricing: Microsoft states "Price based on the country or region in which the exam is proctored" and directs you to confirm with the exam provider at checkout. There is no single global fee published on the AI-901 page; the commonly cited $99 USD is inherited from AI-900 third-party reporting, not stated by Microsoft for AI-901.
- Registration: Use a personal Microsoft account so your records stay portable.
What Microsoft does not publish on the AI-901 page: a fixed question count and a fixed time limit. Figures like "40–60 questions" and "45 minutes" are inherited from AI-900 and may not match the generally available AI-901 form. Confirm format details when you schedule.
Where AI-901 Fits in the Certification Roadmap
AI-901 is a fundamentals exam — an on-ramp, not a prerequisite. It is not mandatory before pursuing associate-level Azure AI certifications, but it is useful if you plan to work toward role-based credentials. Microsoft's current AI credential direction emphasizes Foundry, agentic AI, and generative AI, so AI-901 aligns more closely with the associate-level Azure AI App and Agent Developer (AI-103) path than AI-900 did. If your goal is a career building AI apps on Azure rather than just describing AI concepts, AI-901 is the better starting point. For a broader view of where Azure AI Fundamentals sits among Microsoft fundamentals, see our AZ-900 vs AI-900 vs SC-900 comparison.
Which Exam Should You Choose in 2026?
Because AI-900 retired June 30, 2026, the decision is simpler than competing pages (some written before the retirement) suggest.
Take AI-901 if:
- You are not yet certified and want the Azure AI Fundamentals credential. AI-901 is the only path.
- You are starting fresh and want skills aligned with Microsoft's current AI direction (Foundry, agents, generative AI).
- You plan to pursue the associate-level Azure AI App and Agent Developer (AI-103) certification and want the on-ramp.
You do not need to take either exam again if:
- You already hold the Azure AI Fundamentals certification from AI-900. Your credential is valid permanently. No retake is required.
Consider AI-901 optional if:
- You hold AI-900 and want to demonstrate Foundry and agentic AI skills to employers — a career choice, not a compliance requirement.
You cannot take AI-900 even if you wanted to — it is retired and no longer schedulable as of July 2026. Any "finish AI-900 before June 30" advice you read on older competitor pages is now moot.
Next Steps
- Read the AI-901 study guide for the verified skills measured.
- Check the AI-901 exam page for format, pricing in your region, and scheduling.
- Drill free Azure AI Fundamentals practice questions to test Domain 1 concepts and Foundry implementation under time pressure.
- For the full AI-901 skills breakdown, format details, and a four-week study plan, read our Azure AI Fundamentals AI-901 exam guide.
- For legacy AI-900 content that still helps with Domain 1, see our Azure AI Fundamentals AI-900 exam guide.
AI-901 rewards a different kind of preparation than AI-900 did. Conceptual reading earns you Domain 1 (40–45%). Hands-on Foundry work and Python literacy earn you Domain 2 (55–60%). Treat the exam as a build-and-implement test, not a describe-and-recognize test, and your prep becomes focused instead of scattered.
Buy on Amazon