Microsoft certification practice
Practise Microsoft Azure AI Fundamentals AI-900 practice test — original exam-style scenarios covering every exam domain, with detailed explanations, wrong-answer analysis, and common exam traps. Microsoft certification exams often test whether you can choose the right Azure, Microsoft 365, identity, security or monitoring action for a realistic administrative scenario. Courseiva helps you practise with structured questions and explanations.
New here? Start with this
Recommended: 10-question diagnostic
Covers all 5 domains · takes ~5 min · no account needed · shows your weak areas immediately
Key facts for the Microsoft Azure AI Fundamentals AI-900 exam.
Exam cost
$165 USD / ~£135
Duration
45 minutes
Question count
40–60 questions
Passing score
700 / 1000
Validity
Lifetime (no expiry)
Delivered by
Pearson VUE
Online proctored
Yes — available
Retake policy
100% free — no account required
Everything on Courseiva for AI-900 is free — no subscription, no trial, no email wall.
Courseiva provides free Microsoft Azure AI Fundamentals AI-900 (AI-900) practice questions with original exam-style questions, detailed explanations, topic-based practice, mock exams, readiness tracking, and study analytics. Courseiva is an independent platform and is not affiliated with Microsoft.
24 hours after first failure; 14 days after second
Question formats
Microsoft Azure AI Fundamentals (AI-900) validates foundational understanding of AI and machine learning concepts and how they are implemented in Azure AI services. It is non-technical and suitable for anyone curious about Azure AI regardless of background.
Official prerequisite
None
Recommended experience
No prerequisites. AZ-900 knowledge helpful but not required.
No technical prerequisites. Basic understanding of what AI is used for is helpful.
Should you skip this exam?
If you are a developer or data scientist building AI solutions, go directly to AI-102 (Azure AI Engineer) for meaningful career impact.
Courseiva study tip
Know the difference between classification, regression, clustering, and object detection tasks. The Azure AI services (Azure OpenAI, Azure ML, Cognitive Services) are tested at a conceptual level.
UK salary range
£30,000–£55,000
US salary range
$50,000–$85,000
Job roles that cite AI-900
What certification to take next
$165 USD.
700 out of 1000.
No. AI-900 is a lifetime certification with no expiry.
AI-900 is one of Microsoft's easiest certifications. It tests conceptual awareness of AI and Azure AI services, not implementation skills. Most candidates pass with 15–30 hours of study.
AI-900 is a low-effort credential that demonstrates Azure AI awareness. It adds value as a complement to AZ-900 or AZ-104, but standalone it has limited hiring impact. More valuable for non-technical roles in AI-adjacent organisations.
See how AI-900 stacks up against related certifications.
Practice modes
Choose a mode that matches your current study phase. Use topic practice early on and mock exams closer to your exam date.
All AI-900 Questions
Browse the full question bank organised by domain and difficulty.
Practice Test
Start a session with detailed answer explanations after each question.
Mock Exam
Simulate the real exam with a timed, full-length question set.
Quick 10
A 10-question warm-up — ideal for a short daily study session.
Flashcards
Reinforce key terms, commands, and protocol facts with flashcards.
Domain Practice
Focus exclusively on one exam domain to target your weakest area.
50-Question Set
Mid-length practice set covering all domains.
Full-length Simulation
Full exam simulation matching the real question count.
Exam guide
Start with a short practice session, review each missed answer, then return to the topics that caused mistakes.
Cloud concepts questions usually test the service model (IaaS/PaaS/SaaS) and deployment model (public/private/hybrid/community) appropriate for a given scenario.
IaaS, PaaS and SaaS responsibilities and examples.
Public, private, hybrid and community cloud deployment models.
On-premises vs cloud trade-offs: cost, control, scalability.
How cloud connectivity options (VPN, Direct Connect, ExpressRoute) work.
Related practice questions
Use these internal links to move from the exam guide into focused topic practice pages.
Practise AI-900 questions linked to Describe Artificial Intelligence workloads and considerations.
Practise AI-900 questions linked to Describe fundamental principles of machine learning on Azure.
Practise AI-900 questions linked to Describe features of computer vision workloads on Azure.
Practise AI-900 questions linked to Describe features of Natural Language Processing workloads on Azure.
Practise AI-900 questions linked to Describe features of generative AI workloads on Azure.
Practise AI-900 questions linked to AI-900 fundamentals.
Practise AI-900 questions linked to AI-900 scenario.
Practise AI-900 questions linked to AI-900 troubleshooting.
Practice directory
Every practice set is a unique selection of questions — numbered tests, domain-focused sessions, and topic-specific quizzes.
Showing 30 of 66 tests. See all 66 AI-900 practice tests →
Timed mode
Official blueprint
Target weak areas by practising questions from a single domain. Each domain page shows all questions, difficulty breakdown, and a focused practice session.
Practise AI-900 questions in this domain →
Practise AI-900 questions in this domain →
Practise AI-900 questions in this domain →
Practise AI-900 questions in this domain →
Practise AI-900 questions in this domain →
Term reference
These terms appear frequently in AI-900 questions. Understanding them precisely reduces guessing and improves scores.
Azure AI Language
Azure AI Language is a cloud-based service from Microsoft that uses natural language processing to understand, analyze, and generate human language for applications.
Azure AI Search
Azure AI Search is a cloud search service that uses artificial intelligence to index and search through large amounts of data, returning relevant results that feel as smart as a Google search.
Azure AI Services
Azure AI Services is a collection of pre-built, cloud-based artificial intelligence APIs and services that allow developers and IT professionals to integrate capabilities like vision, speech, language, and decision-making into applications without needing deep machine learning expertise.
Azure AI Speech
Azure AI Speech is a cloud service from Microsoft that converts spoken audio into text, text into lifelike speech, and enables real-time voice translation and speaker recognition.
Azure AI Translator
Azure AI Translator is a cloud-based service that uses neural machine translation to convert text between languages while preserving meaning and context.
Azure AI Vision
Azure AI Vision is a cloud-based service from Microsoft that uses pre-built machine learning models to extract information from images and videos, such as objects, text, faces, and scene descriptions.
Azure Machine Learning
Azure Machine Learning is a cloud service for building, training, and deploying machine learning models at scale.
Azure OpenAI Service
Azure OpenAI Service is a cloud platform from Microsoft that lets developers use powerful artificial intelligence models, like GPT-4, to build applications that can understand and generate human-like text, code, images, and more.
Free study guide
The Courseiva AI-900 study guide covers exam format, the official domain blueprint with weights, a structured study plan, and domain-specific tips for each section. Use it alongside practice questions to build both conceptual depth and exam confidence.
Sample questions
Select your answer before revealing the explanation. These questions are original exam-style — not copied from any exam or dump site.
Trap 1: Fairness
Fairness addresses bias and discrimination; while lack of transparency could hide bias, the core violation here is the inability to explain the decision.
Trap 2: Privacy and security
Privacy and security focus on data protection, not on providing explanations for decisions.
Trap 3: Reliability and safety
Reliability and safety ensure the system operates correctly and without harm, but they do not directly require explainability.
Fairness
Why wrong: Fairness addresses bias and discrimination; while lack of transparency could hide bias, the core violation here is the inability to explain the decision.
Transparency
Transparency requires that systems are explainable and that users are informed about how decisions are made.
Privacy and security
Why wrong: Privacy and security focus on data protection, not on providing explanations for decisions.
Reliability and safety
Why wrong: Reliability and safety ensure the system operates correctly and without harm, but they do not directly require explainability.
Trap 1: Updating the Python version used to run ML training scripts
Python version management is environment configuration — model versioning tracks trained model artefacts and their lineage.
Trap 2: Releasing new features of the Azure ML service as versioned API…
Azure service versioning is product development — model versioning is the practice of tracking your own trained model iterations.
Trap 3: Managing multiple versions of training data used by different model…
Data versioning is a related but separate practice — model versioning specifically tracks the trained model artefacts.
Updating the Python version used to run ML training scripts
Why wrong: Python version management is environment configuration — model versioning tracks trained model artefacts and their lineage.
Tracking each iteration of a trained model for rollback, A/B testing, auditing, and reproducibility
Model versioning enables safe updates (rollback when new version fails), experimentation (A/B test), and regulatory compliance (audit trail).
Releasing new features of the Azure ML service as versioned API updates
Why wrong: Azure service versioning is product development — model versioning is the practice of tracking your own trained model iterations.
Managing multiple versions of training data used by different model experiments
Why wrong: Data versioning is a related but separate practice — model versioning specifically tracks the trained model artefacts.
Trap 1: Making AI model source code publicly available as open source
Open-source code is one form of transparency — AI transparency broadly covers disclosure of AI use, data, limitations, and decision rationale.
Trap 2: Reporting all AI project costs transparently in financial statements
Financial reporting is accounting — AI transparency is about understanding AI systems' behaviour, data, and limitations.
Trap 3: Making all training data publicly available for independent…
Training data disclosure is one transparency measure — AI transparency broadly applies to all aspects of how AI systems work and are used.
Making AI model source code publicly available as open source
Why wrong: Open-source code is one form of transparency — AI transparency broadly covers disclosure of AI use, data, limitations, and decision rationale.
Ensuring people understand when they're interacting with AI, how it works, and what its limitations are
Transparency covers AI disclosure, model documentation, explainability, and honest uncertainty communication — building informed trust.
Reporting all AI project costs transparently in financial statements
Why wrong: Financial reporting is accounting — AI transparency is about understanding AI systems' behaviour, data, and limitations.
Making all training data publicly available for independent researchers to audit
Why wrong: Training data disclosure is one transparency measure — AI transparency broadly applies to all aspects of how AI systems work and are used.
Trap 1: Decrease the temperature parameter.
Decreasing temperature makes the model more deterministic and less random, but it does not specifically target repetition; it may reduce creativity and still produce loops.
Trap 2: Increase the presence_penalty parameter.
Increasing presence_penalty penalizes any token that has appeared at all, which can reduce topic repetition but frequency_penalty is more precise for repetitive phrases.
Trap 3: Decrease the top_p parameter.
Decreasing top_p makes the model consider only a smaller set of likely tokens, which can make outputs more focused but does not directly address repetition of already used tokens.
Increase the frequency_penalty parameter.
Correct. Increasing frequency_penalty reduces the likelihood that the model will repeat the same tokens, making it effective against repetitive loops.
Decrease the temperature parameter.
Why wrong: Decreasing temperature makes the model more deterministic and less random, but it does not specifically target repetition; it may reduce creativity and still produce loops.
Increase the presence_penalty parameter.
Why wrong: Increasing presence_penalty penalizes any token that has appeared at all, which can reduce topic repetition but frequency_penalty is more precise for repetitive phrases.
Decrease the top_p parameter.
Why wrong: Decreasing top_p makes the model consider only a smaller set of likely tokens, which can make outputs more focused but does not directly address repetition of already used tokens.
Trap 1: Accuracy
Accuracy is a classification metric that measures the proportion of correct predictions, not suitable for continuous output.
Trap 2: F1 Score
F1 score is a classification metric that balances precision and recall, not used for regression.
Trap 3: Precision
Precision is a classification metric that measures the proportion of true positives among positive predictions, not applicable to regression.
Root Mean Squared Error (RMSE)
RMSE measures the average magnitude of prediction errors in the original units, making it suitable for regression.
Accuracy
Why wrong: Accuracy is a classification metric that measures the proportion of correct predictions, not suitable for continuous output.
F1 Score
Why wrong: F1 score is a classification metric that balances precision and recall, not used for regression.
Precision
Why wrong: Precision is a classification metric that measures the proportion of true positives among positive predictions, not applicable to regression.
Trap 1: Prompt engineering
Prompt engineering involves crafting the input prompt to guide the model, but it cannot guarantee that the model will avoid generating harmful content on its own.
Trap 2: Fine-tuning
Fine-tuning adapts the model to a specific style or domain but does not automatically filter harmful content from the output.
Trap 3: Few-shot learning
Few-shot learning provides examples in the prompt to set context, but it does not provide a safety filter against offensive outputs.
Content filtering
Content filtering applies safety rules to block offensive or harmful language in model outputs, regardless of the prompt's phrasing.
Prompt engineering
Why wrong: Prompt engineering involves crafting the input prompt to guide the model, but it cannot guarantee that the model will avoid generating harmful content on its own.
Fine-tuning
Why wrong: Fine-tuning adapts the model to a specific style or domain but does not automatically filter harmful content from the output.
Few-shot learning
Why wrong: Few-shot learning provides examples in the prompt to set context, but it does not provide a safety filter against offensive outputs.
Official booking and top-rated study resources for the AI-900.
Question Discussion
Share a tip, memory trick, or ask about the reasoning behind this question. Do not post real exam questions, leaked content, braindumps, or copyrighted exam material. Comments are moderated and may be removed without notice.
Sign in to join the discussion.