AI-901 Exam Guide: How to Pass Azure AI Fundamentals in 2026

Last updated: 22 April 2026

Microsoft's brand-new Exam AI-901: Microsoft Azure AI Fundamentals opened for beta registration yesterday, replacing the long-serving AI-900 exam that retires on 30 June 2026. If you are a UK developer, analyst, student, or career changer looking for an entry-level Microsoft AI certification, AI-901 is now the exam you will sit to earn the Azure AI Fundamentals badge.

This guide walks you through everything you need to pass AI-901 first time: what the exam covers, what Microsoft Foundry is and why it sits at the heart of the syllabus, a week-by-week study plan, the free Microsoft Learn modules that mirror the skills outline, and a realistic difficulty assessment. It is written for UK candidates and reflects the official skills measured as of 15 April 2026.

Early-mover advantage: AI-901 is live in beta now. The first 300 candidates to sit the exam on or before 6 May 2026 can use Microsoft's beta promo code AI901Medford for 80% off the standard price. Details in the pricing section below.

What Is the AI-901 Exam?

AI-901, formally titled Exam AI-901: Microsoft Azure AI Fundamentals, is the replacement exam for Microsoft's retiring AI-900 certification. Passing it earns the same credential as before: Microsoft Certified: Azure AI Fundamentals. Microsoft chose to update the underlying exam rather than change the certification name, which keeps existing AI-900 badges valid and avoids fragmenting the fundamentals career path.

The headline difference is platform focus. AI-900 asked candidates to describe what a catalogue of Azure AI services could do. AI-901 asks you to implement AI solutions using a single, unified platform: Microsoft Foundry. In Microsoft's own words:

The updated Microsoft Certification validates your knowledge of AI concepts and capabilities and your ability to implement AI solutions by using Azure and Microsoft Foundry." --Microsoft Tech Community, 15 April 2026

In practical terms, that shift changes who this exam is right for, what you need to study, and how hands-on the preparation has to be. We unpack all three below.

Who Should Take AI-901 in 2026?

AI-901 is explicitly pitched at technical beginners who plan to build AI solutions. It is a meaningful shift from AI-900, which served a much broader non-technical audience.

According to the official AI-901 exam page:

This certification is intended for individuals who want to start working with AI solutions built on Azure. It is suitable for learners from technical backgrounds, including aspiring junior developers who are starting to incorporate AI capabilities into applications.

AI-901 is a good fit if you are:

A junior developer starting to add AI capabilities to your applications
A data or business analyst moving into AI-adjacent work in Microsoft Foundry
A computer science or informatics student looking for a recognised Microsoft credential before entering the job market
A career changer transitioning from another IT role into AI engineering
Someone holding AI-900 who wants to refresh their credential on the current platform

It is probably not the right first exam for business stakeholders who will never write code, purely operational AI users (such as Microsoft 365 Copilot end users), or architects looking for a senior-level AI certification. For those audiences, the AB-730 and AB-731 pathways remain more appropriate.

Microsoft recommends candidates arrive with:

Familiarity with basic cloud concepts such as storage, compute, and authentication
Basic Python syntax and programming technique
A free Azure account for hands-on practice

If you sat AI-900 and are now weighing whether to take AI-901 before the old exam retires, read our dedicated AI-900 vs AI-901 transition guide first.

AI-901 Exam Format at a Glance

Here is the confirmed format for AI-901 as at 22 April 2026. Microsoft has not yet published every logistical detail during the beta window, and the table flags the gaps honestly.

Attribute	Value
Exam code	AI-901
Official title	Microsoft Azure AI Fundamentals
Certification earned	Microsoft Certified: Azure AI Fundamentals
Current status	Beta (opened 21 April 2026)
Passing score	700 / 1,000
Number of skill domains	2
Delivery	Pearson VUE test centre or OnVUE online proctored
Languages	13 (English and major world languages)
Formal prerequisites	None
Recommended background	Basic cloud concepts, basic Python
Exam duration	To be confirmed for GA
Number of questions	To be confirmed for GA
Free Microsoft practice assessment	Not yet available; typically released around 8 weeks after GA

A few logistical notes worth flagging:

Beta vs GA: The beta sitting is the same exam content as GA, but your score is released approximately 10 days after the beta window closes. You can still earn the Azure AI Fundamentals certification from a beta pass.
Duration and question count are not yet publicly stated by Microsoft for AI-901 specifically. The retiring AI-900 sat around 45 minutes with roughly 40 to 60 questions. AI-901 is expected to be similar, but always verify the current details on the official exam page when you book.
Certification validity: Fundamentals certifications do not expire, so a pass in 2026 remains on your transcript indefinitely.

AI-901 Skill Domains and Weighting Explained

The AI-901 syllabus is divided into two skill domains, confirmed by Microsoft's official study guide published on 15 April 2026.

Domain 1: Identify AI concepts and responsibilities (40 to 45%)

Domain 1 is the theory half of the exam. It covers responsible AI principles, how generative AI models work at a conceptual level, and how to recognise common AI workloads.

Microsoft lists the following sub-skills:

Describe principles of responsible AI

Describe considerations for fairness in an AI solution
Describe considerations for reliability and safety in an AI solution
Describe considerations for privacy and security in an AI solution
Describe considerations for inclusiveness in an AI solution
Describe considerations for transparency in an AI solution
Describe considerations for accountability in an AI solution

Identify AI model components and configurations

Describe how generative AI models work
Identify an appropriate AI model, based on capabilities
Identify appropriate model deployment options and configuration parameters

Identify AI workloads

Identify scenarios for common AI workloads, including generative and agentic AI, text analysis, speech, computer vision, and information extraction
Describe common text analysis techniques, including keyword extraction, entity detection, sentiment analysis, and summarisation
Identify features and capabilities of speech recognition and speech synthesis
Identify features and capabilities of computer vision and image-generation models
Identify techniques to extract information from text, images, audio, and videos

Domain 2: Implement AI solutions by using Microsoft Foundry (55 to 60%)

Domain 2 is the larger half and it is hands-on. Every sub-skill expects you to build, deploy, or interact with something in the Foundry portal or SDK. This is the single biggest behavioural change from AI-900: passive recognition has been replaced by applied implementation.

Microsoft groups the Domain 2 sub-skills into four clusters.

Implement generative AI apps and agents by using Foundry

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

Implement AI solutions for text and speech by using Foundry

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

Implement AI solutions with computer vision and image-generation capabilities by using Foundry

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

Implement AI solutions for information extraction by using Foundry

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

If you count the sub-skills, Domain 2 is both broader and more practical than Domain 1. Budget your preparation time accordingly: roughly 60% of your study should go into Foundry-based exercises, 40% into responsible AI and workload recognition.

Microsoft Foundry: The Biggest Change from AI-900

You cannot pass AI-901 without being comfortable inside Microsoft Foundry. Microsoft's Foundry overview page defines it concisely:

Microsoft Foundry is a unified Azure platform-as-a-service offering for enterprise AI operations, model builders, and application development. This foundation combines production-grade infrastructure with friendly interfaces, enabling developers to focus on building applications rather than managing infrastructure.

In practice, Foundry folds together several previously separate products:

Before	Now
Azure AI Studio / Azure AI Foundry (brand)	Microsoft Foundry
Azure AI Services (brand)	Foundry Tools
Assistants API (Agents v0.5 / v1)	Responses API (Agents v2)
Hub + Azure OpenAI + Azure AI Services	Single Foundry resource with projects
Multiple SDKs against 5+ endpoints	Unified project client, one project endpoint

Three implications for candidates:

Hands-on time in the Foundry portal at ai.azure.com is non-negotiable. Reading the docs is not enough, because Domain 2 asks you to deploy models, test prompts, and build chat and agent apps. You need to have done those things at least once.
Terminology has moved on from AI-900. Do not rely on old study notes that refer to Azure AI Studio, Assistants API, or the old hub-plus-services pattern. The exam uses current Foundry vocabulary.
The SDK shift matters. AI-901 includes "create a lightweight chat client application by using the Foundry SDK", so you need to install the Foundry client and make at least one working call before exam day.

A useful grounding read before you start the modules is the Foundry overview documentation. Read it end to end once, then revisit individual capability pages as you work through Domain 2.

Is AI-901 Hard? A Realistic Difficulty Assessment

Honest answer: AI-901 is harder than AI-900, even for the same audience. Three structural changes drive the step up.

Change 1: It is now hands-on. AI-900 could be passed largely from reading and watching videos. AI-901 expects you to have deployed a model, sent a chat completion, tested an agent, and run Content Understanding against a document. Candidates who try to read their way through will be punished in the larger Domain 2.

Change 2: Python literacy is now required. Microsoft states AI-901 candidates should have basic Python syntax and programming technique. You do not need to be a production Python engineer, but you need to read and run short scripts, understand what a function call looks like, and feel comfortable with pip install.

Change 3: Responsible AI went deeper. AI-900 treated responsible AI as a single domain bullet. AI-901 breaks it into six principles (fairness, reliability and safety, privacy and security, inclusiveness, transparency, accountability) and asks you to apply considerations rather than simply describe them. Expect scenario questions, not just definitions.

That said, AI-901 is still a fundamentals exam, not an engineer exam. It is graded at 700 out of 1,000 (a standard Microsoft fundamentals threshold), there are no formal prerequisites, and the official one-day course (AI-901T00-A) confirms the scope is intentionally introductory. Most UK candidates with a modest technical background should be able to prepare in three to four weeks of focused study.

We are not guaranteeing a pass and nobody honest can. What we will say is: treat Domain 2 as a practical skill to build, not a reading exercise, and your chances climb significantly.

The 4-Week AI-901 Study Plan

This plan assumes six to eight hours per week of focused study and access to a free Azure account for hands-on work in Foundry. Adjust down to three weeks if you already have Azure experience, or up to six weeks if you are new to both cloud and Python.

Week 1: AI concepts and responsible AI (Domain 1 focus)

Goals: build a clear mental model of what AI is, how generative models work, and how Microsoft expects you to reason about responsible AI.

Read the AI-901 study guide end to end
Complete Module 1 of the Microsoft Learn path: Get started with AI in Azure
Read and make notes on each of the six responsible AI principles, writing your own one-sentence example of each
Read the Microsoft Foundry overview to set up for Week 2

End of week checkpoint: you can explain fairness, reliability and safety, privacy and security, inclusiveness, transparency, and accountability in your own words, with a realistic scenario for each.

Week 2: Foundry fundamentals and generative AI (Domain 2, cluster 1)

Goals: get your hands dirty in the Foundry portal and build your first generative AI applications.

Create a free Azure account if you do not already have one
Complete Module 2: Get started with generative AI and agents in Azure
Deploy a model in the Foundry portal at ai.azure.com and test a prompt in the chat playground
Install the Foundry SDK and run a minimal chat client (Microsoft's tutorial in the module walks you through this)
Create and test one single-agent solution in the portal, even a trivial one

End of week checkpoint: you have deployed at least one model, sent chat completions both in the portal and via the SDK, and built a simple agent.

Week 3: Text, speech, vision, and information extraction (Domain 2, clusters 2 to 4)

Goals: cover the rest of Domain 2's applied clusters. This is the heaviest week.

Complete Module 3: Get started with text analysis in Azure
Complete Module 4: Get started with speech in Azure
Complete Module 5: Get started with computer vision in Azure
Complete Module 6: Get started with AI-powered information extraction in Azure
Run at least one Content Understanding job against a PDF or image

End of week checkpoint: you have built at least a lightweight app in each of text, speech, vision, and information extraction.

Ready to test yourself on AI-901? Examinotion's AI-901 practice exam gives you 40-question, exam-format tests with detailed explanations for every answer. Use one after Week 3 to benchmark your readiness before final revision.

Week 4: Practice tests, weak-area revision, and exam preparation

Goals: benchmark your readiness, revise weak areas, and get familiar with the exam interface.

Take a full-length AI-901 practice exam under timed conditions
For every question you get wrong, go back to the corresponding Microsoft Learn module and re-read the section
Revisit the responsible AI principles one more time (they are easy marks)
Explore the Microsoft exam sandbox so the exam-day interface is not new to you
Book your exam slot (see booking section below)

End of week checkpoint: you are scoring at least 80% on practice exams, you know which responsible AI principle applies to typical scenarios, and you have an exam slot in the calendar.

Essential AI-901 Topics You Must Master

A few topics sit at the intersection of multiple sub-skills and deserve extra attention.

Responsible AI principles applied to scenarios. Do not memorise the six principles in isolation. For each principle, be able to state one concrete example of a violation (fairness: a hiring model that scores candidates differently by gender), one mitigation (apply a bias assessment before deployment), and one design artefact (a datasheet or a transparency note).

System prompts vs user prompts. Know the structural difference and when to set each. The exam explicitly asks you to create effective system and user prompts for generative AI models.

Model deployment options. Understand serverless deployment versus dedicated endpoints, when each makes sense, and the key configuration parameters (temperature, top-p, max tokens, frequency and presence penalties).

Single-agent versus multi-agent. AI-901 only requires single-agent understanding, but you should know what an agent is, how it uses tools and memory, and how Responses API (Agents v2) differs from the older Assistants API.

Multimodal models. The exam explicitly includes responding to spoken prompts using a multimodal model and interpreting visual inputs in prompts. You need to have used one, not just read about it.

Content Understanding. The information extraction cluster centres on Azure Content Understanding in Foundry Tools. Run it against a PDF, an image, and an audio file before exam day so the SDK shape is familiar.

Best Free and Paid AI-901 Study Resources (UK)

All the resources below are current as at 22 April 2026.

Free Microsoft resources

Official AI-901 exam page
Official AI-901 study guide
Learning path: Get started with AI applications and agents on Azure (6 modules)
Microsoft Foundry documentation
Microsoft exam sandbox (exam interface preview)

Paid training

AI-901T00-A: Introduction to AI in Azure (one-day instructor-led course; typically booked through a Microsoft training partner)

Paid practice tests

Examinotion AI-901 practice exam: 40-question tests matching the real AI-901 format, with detailed explanations for every answer, at £12.99 for unlimited access until you pass.

The official Microsoft free practice assessment is not yet available for AI-901. Microsoft typically releases those assessments around eight weeks after a beta exam reaches general availability.

How to Book and Sit the AI-901 Exam in the UK

AI-901 is delivered through Pearson VUE, either at a test centre or via OnVUE online proctoring from home. The booking flow is the same one used for every Microsoft certification.

Step-by-step booking:

Sign in to your Microsoft Learn profile with the Microsoft account you want the credential associated with.
Go to the AI-901 exam page and select Schedule exam.
Choose Pearson VUE as the delivery provider.
Pick either an in-person test centre or OnVUE online proctored.
Select a date and time. Exams can be scheduled up to 90 days in advance.
Pay for the exam. If you are using the beta promo code, enter AI901Medford at checkout.

UK candidate notes:

Test centres are widely available in London, Manchester, Birmingham, Edinburgh, Glasgow, Cardiff, and most larger cities. Availability varies, so book early if you have a preferred date.
OnVUE requires a quiet space, a working webcam and microphone, a government ID, and a stable internet connection. Check the Pearson VUE system test before exam day.
Pearson VUE UK customer service is on freephone 0800-7319905, Monday to Friday, 09:00 to 18:00 CET.

Pricing

Microsoft does not publish a single global price for AI-901. The official exam page states that price is based on the country or region where the exam is proctored. The actual GBP figure is shown during the Pearson VUE scheduling flow after you pick a UK delivery option, so verify your current price there before paying.

Beta promo (time-sensitive, confirmed by Microsoft):

Code: AI901Medford
Discount: 80% off the standard price
Eligibility: the first 300 candidates globally to sit the beta on or before 6 May 2026
Not valid in Turkey, Pakistan, India, or China

If you are reading this before 6 May 2026 and you are confident about your preparation, the beta discount is the cheapest way to earn the Azure AI Fundamentals credential in 2026.

Separately, Examinotion's AI-901 practice exam pack is £12.99, fixed, with unlimited attempts until you pass your real exam.

10 AI-901 Exam-Day Tips to Maximise Your Score

Run the Pearson VUE system test at least 24 hours before your OnVUE sitting. Fixing a webcam driver five minutes before start time is not a strategy.
Read every question twice before looking at the answers. Microsoft questions are precisely worded and small qualifiers change the correct option.
Answer the easy questions first and flag anything you are unsure about. Come back at the end with a clearer head.
Use the review screen at the end to revisit flagged questions. Do not run down the clock on one hard question early.
Do not second-guess responsible AI questions. The six principles map cleanly to most scenarios. Pick the best fit and move on.
For implementation questions, picture the Foundry portal. If you have actually deployed a model and tested a prompt, the correct answer is usually obvious.
Eliminate obviously wrong options first. Even when you are unsure, narrowing four options to two doubles your odds.
Trust "none of the above" only when the other options are provably wrong. It is rarely the right answer on Microsoft exams.
Do not leave any question blank. There is no penalty for wrong answers on Microsoft fundamentals exams.
Stay off Reddit and braindump forums. They are at best distractions and at worst copyright and NDA violations that can invalidate your result.

Frequently Asked Questions

How hard is the AI-901 exam?

AI-901 is a step up from AI-900 because Domain 2 expects hands-on Microsoft Foundry work and the exam now assumes basic Python literacy. That said, it remains a fundamentals-tier exam with no formal prerequisites, a standard 700/1,000 pass mark, and roughly 55 to 60% of the weight on practical implementation. Most UK candidates with modest technical experience can prepare in three to four weeks of focused study.

How long should I study for AI-901?

Plan for three to four weeks at six to eight hours per week if you are new to the content. That covers the two-domain syllabus, all six Microsoft Learn modules in the Foundry learning path, hands-on Foundry portal time, and one full-length practice exam before booking. Stretch to five or six weeks if you are also new to Python or the Azure portal.

What is the pass score for AI-901?

The pass score for AI-901 is 700 out of 1,000, the standard threshold for Microsoft fundamentals exams. Exam scores are released approximately 10 days after the end of the beta window for candidates sitting the beta, and immediately on completion once the exam reaches general availability.

Is AI-901 worth it in 2026?

Yes, for the right audience. AI-901 is currently the entry-level Microsoft AI credential for technical learners from 21 April 2026 onward, it earns the same Microsoft Certified: Azure AI Fundamentals badge as the retiring AI-900, and Microsoft Foundry is now the platform underpinning most of Microsoft's AI stack. If you are a junior developer or career changer adding AI to your CV in 2026, AI-901 is the cheapest, fastest-recognised proof point.

Can I pass AI-901 without Azure experience?

Technically yes, because there are no formal prerequisites, but it is much harder. Domain 2 is 55 to 60% of the exam and it is almost entirely hands-on work in Microsoft Foundry. Realistically, you should plan at least one week of Foundry portal and SDK practice before exam day. A free Azure account is enough; you do not need a paid subscription.

Do I need to take AI-901 if I already hold AI-900?

No. Microsoft confirms that Fundamentals certifications do not expire, so your AI-900 credential remains valid indefinitely. Many candidates choose to re-earn Azure AI Fundamentals via AI-901 to signal currency with Microsoft Foundry, but it is not required. If you are considering whether to take AI-900 before it retires or jump straight to AI-901, read our AI-900 vs AI-901 transition guide for a scenario-based recommendation.

Start Your AI-901 Preparation Today

AI-901 is live in beta, AI-900 retires on 30 June 2026, and the first wave of UK candidates has a real opportunity: the exam is new enough that competing preparation resources are still being written, and the syllabus is stable enough to study confidently against Microsoft's published skills outline.

Pair this guide with the free Microsoft Learn path and the official study guide. When you are ready to test yourself under timed, exam-format conditions, Examinotion's AI-901 practice exam gives you 40-question tests with explanations for every answer at £12.99, unlimited attempts until you pass.

Good luck on exam day.

Sources: