Azure DP-900 vs AI-900: Which Should You Take First?

Forget everything Reddit told you about Azure fundamentals exams being "pointless." For a lot of people starting out, DP-900 or AI-900 is the smartest first move you can make — but only if you pick the right one for where you're headed.

I took both back-to-back over two weekends. One felt like a comfortable review. The other genuinely surprised me with how much I didn't know. Let me save you from making the same mistake I almost did.

The Quick Comparison (for Impatient People)

What DP-900 Actually Covers

The Azure Data Fundamentals (DP-900) exam is Microsoft's way of testing whether you understand core data concepts. And honestly, a lot of it is stuff you might already know if you've ever worked with databases.

The Four Domains

• Core data concepts (25-30%): Types of data (structured, semi-structured, unstructured), data storage models, data processing (batch vs streaming). Pretty straightforward stuff.
• Relational data on Azure (20-25%): Azure SQL Database, Azure SQL Managed Instance, PostgreSQL, MySQL. If you've used any SQL database, this feels familiar.
• Non-relational data on Azure (15-20%): Cosmos DB, Azure Blob Storage, Azure Table Storage. This is where it gets more Azure-specific.
• Analytics workloads on Azure (25-30%): Azure Synapse, Azure Data Factory, Power BI. The analytics section carries significant weight.

Who Finds DP-900 Easy

If you've ever written a SQL query, built a dashboard, or worked with an ETL pipeline, you'll find at least 60% of DP-900 is review. The Azure-specific services are the main learning curve.

Developers, database admins, and anyone who's touched Power BI will breeze through most of it.

What AI-900 Actually Covers

The Azure AI Fundamentals (AI-900) exam tests your understanding of artificial intelligence and machine learning concepts, plus Azure's AI service offerings.

The Five Domains

• AI workloads and considerations (15-20%): What AI is, responsible AI principles. The ethics stuff catches people off guard.
• Machine learning fundamentals (20-25%): Supervised vs unsupervised learning, regression, classification, clustering. The theory section.
• Computer vision workloads (15-20%): Azure Computer Vision, Custom Vision, Face API. Real-world image analysis.
• NLP workloads (15-20%): Azure Language Understanding, text analytics, speech services, Azure OpenAI. This got updated significantly in 2025-2026.
• Generative AI workloads (15-20%): Azure OpenAI Service, prompt engineering basics, responsible generative AI. The newest section and the one everyone underestimates.

Who Finds AI-900 Tricky

If you've never studied machine learning, the ML fundamentals section can feel abstract. Concepts like precision vs recall, overfitting, and feature engineering aren't intuitive to most IT professionals. And the responsible AI section? Microsoft takes this very seriously — expect 5-8 questions specifically about ethics and fairness.

DP-900 vs AI-900: Head-to-Head Breakdown

Difficulty Comparison

DP-900 is slightly easier for most people. Why? Because data concepts are more tangible. Everyone understands what a table is, what a database does, what "filtering data" means. AI concepts like neural networks, NLP pipelines, and confusion matrices are inherently more abstract.

That said, if you have a data science or ML background, you'll find AI-900 trivially easy. It really depends on your starting point.

Career Path Comparison

This is where the decision actually matters.

DP-900 leads to:

• Azure DP-300 (Database Administrator) — solid, stable career path
• DP-203 (Data Engineer) — one of the hottest Azure roles right now
• Data analyst roles using Power BI and Azure Synapse

AI-900 leads to:

• Azure AI-102 (AI Engineer) — increasingly in demand post-GPT era
• Azure DP-100 (Data Scientist) — for ML model builders
• AI solution architect roles combining cloud + ML

Content Overlap

There's about 10-15% overlap between the two exams. Both cover basic Azure concepts and some data fundamentals. If you take one, the other becomes slightly easier. But they're different enough that you can't study for one and accidentally pass the other.

Which One Should You Take First?

Here's my decision framework:

Take DP-900 First If...

• You have any SQL or database experience
• You want the easiest win to build confidence
• Your goal is data engineering, BI, or database administration
• You're completely new to Azure and want a gentle introduction
• You already work with data but want to formalize your knowledge

Take AI-900 First If...

• You're genuinely excited about AI/ML (motivation matters!)
• You have Python experience or ML background
• Your goal is AI engineering or data science
• You want to understand Azure OpenAI Service for your current role
• Your company is investing in AI and you need to speak the language

Take Both If...

Honestly? If you can afford the time and $198, take both. They each take 2-3 weeks of casual studying, and having both fundamentals certs on your resume signals breadth. I did them back-to-back and the second exam was notably easier because of the Azure familiarity from the first.

Study Strategy for Each Exam

DP-900 Study Plan (2-3 Weeks)

1. Week 1: Microsoft Learn DP-900 learning path (free!) — covers everything you need
2. Week 2: Practice questions on ExamCert's free DP-900 practice test
3. Week 3 (optional): Review weak areas, do one more full practice test

Key tip: The analytics section (Synapse, Data Factory, Power BI) carries the most weight and is the most Azure-specific. Spend extra time here.

AI-900 Study Plan (2-3 Weeks)

1. Week 1: Microsoft Learn AI-900 learning path + the new Generative AI module
2. Week 2: Practice questions on ExamCert's free AI-900 practice test
3. Week 3 (optional): Deep dive into responsible AI principles and Azure OpenAI concepts

Key tip: Don't skip the responsible AI content. I know it feels fluffy compared to technical questions, but Microsoft dedicates a surprising number of questions to ethics, fairness, transparency, and accountability. It's easy points if you study it, easy fails if you don't.

How These Compare to AZ-900

Since people always ask: AZ-900 (Azure Fundamentals) is the general Azure fundamentals cert. It's broader than both DP-900 and AI-900 but less deep in any specific area.

The recommended order for most people:

1. AZ-900 first (general Azure knowledge)
2. DP-900 or AI-900 second (specialization signal)
3. Role-based cert third (AZ-104, DP-300, or AI-102)

But if you already have cloud experience, skip AZ-900 and go straight to DP-900 or AI-900.

The Bottom Line

Both exams cost $99, take 2-3 weeks to prepare for, and serve as stepping stones to much more valuable role-based certifications. Neither one will land you a job on its own, but both signal to employers that you're serious about your chosen path.

If I had to pick just one? DP-900 for the safer bet, AI-900 for the bigger potential upside. But honestly, the best choice is the one that aligns with what gets you excited. You'll study harder and learn more when you actually care about the content.

Frequently Asked Questions