Microsoft is Retiring the DP-100

Quick Answer: Starting June 1, the DP-100 exam is retired, and in its place is a certification with more focus on machine learning and generative AI. If you’ve been working on your Azure AI certification, you may need to make a few adjustments based on your scope and progress.

Microsoft is retiring DP-100: Azure Data Scientist Associate, effective June 1, 2026. As part of Microsoft's shift to AI-first operations, many of the old certifications will be discontinued. A new MLOps certification, AI-300: MLOps Engineer Associate, will take its place.

Why Microsoft is Retiring DP-100

As AI becomes the norm, Microsoft is keen to overhaul its certifications to accommodate the evolving work environments. The company wants its certs to keep up with modern expectations for data scientists who have a greater need for machine learning and generative AI. 

Microsoft structured DP-100 around traditional data science. But as machine learning operations (MLOps) continue to establish a foothold across many industries, it demands that data scientists have additional skill sets. Modern roles now require IT professionals to be adept at operating and managing, rather than just pure design. In short, data scientists are increasingly taking on more operational roles.

The Azure Machine Learning service was one of the primary platforms for DP-100. However, the new certs will support additional platforms, such as Microsoft Fabric and Databricks.

Notable Changes Between DP-100 vs. AI-300

AI-300 includes several changes in scope and focus. While the primary focus of DP-100 is the design and training of ML models, AI-300 places greater emphasis on operations (MLOps) and operationalizing AI.

Key skills needed to pass DP-100 include model training and evaluation. For AI-300, your expertise should be in automation and governance. AI-300 also places more weight on emerging technologies, such as generative AI. 

The exam style is also different. While the old DP-100 exam is heavy on coding and theory, AI-300 is more architectural.

What the AI-300 Exam Covers

These domains make up the bulk of the new AI-300 exam:

• Designing and deploying MLOps platforms and pipelines

• Managing the end-to-end lifecycle and operational workflows of machine learning models

• Building and configuring GenAIOps environments and tooling

• Establishing QA processes and observability for generative AI applications

• Tuning generative AI systems and improving model performance

Additionally, you are expected to be adept at the following tools and practices:

• Azure Machine Learning

• Microsoft AI Foundry

• GitHub Actions

• IaC Tools

• Python SDK v2

As of the beta, the exam has up to 60 questions, with a time limit of approximately 120 minutes.

What to Do If You’re Mid-Prep for DP-100

If you’re already preparing to take the DP-100 exam, whether to advance or pivot depends on how much training you’ve already covered, as well as your career plans.

You have until June 1 to take and pass the exam. The credential itself will not expire on this date, and will continue to appear on your Microsoft Learn transcript until it expires. 

If you’re already well into learning major domains like Azure Machine Learning and traditional model training, it’s logical to just continue and finish. Check the study guide for more information.

But if you’ve only just started your studies, it’s better to just change course and refocus on AI-300. AI-300 is a future-ready certification for generative AI and MLOps. Potential recruiters will likely look for AI-first credentials moving forward.

What if You Already Have DP-100

No immediate action needed on your part if you already have the DP-100 cert. The certificate itself will remain in your profile until its specific expiration date; only the exam will discontinue soon. Of course, once the certification expires, you won’t be able to renew it. You should therefore plan accordingly. 

What Your Next Step Should Be

The next step depends on your current status and future career plans. If you’re about to finish your DP-100 training, take the exam and retain the certification until its expiration. If you’re not too invested in the training yet, or if you’ve yet to train at all, it makes more sense to proceed to AI-300 instead.

Whether you’re training for machine learning, generative AI, Azure, or MLOps courses, CBT Nuggets offers comprehensive Microsoft training that fits your study program. Keep your skills sharp and future-proof with CBT Nuggets courses. Sign up and your first 7 days are free.

Frequently Asked Questions

1. What is the exact retirement date for DP-100? 

The DP-100: Microsoft Azure Data Scientist Associate exam will be officially retired on June 1, 2026. It is one of many courses Microsoft plans to discontinue. Over the next 12 months, more than a dozen Microsoft exams will be retired.

2. Does my existing DP-100 certification get invalidated when the exam retires?

No, your certification is still valid until its specific expiration. It will remain in the active section of your Microsoft Learn transcript until it expires, typically after a year. After that, it will be placed in the Historical Certifications section of your Microsoft Learn transcript.

3. Can I still renew my DP-100 before it retires?

Yes, it’s possible to renew your DP-100 Azure Data Scientist certification, as long as you’re up for renewal before the retirement date. You must complete and pass the free online renewal assessment on Microsoft Learn before June 1, 2026.

4. Is AI-300 harder than DP-100?

Although not necessarily harder, the AI-300 exam is more technically complex. Data scientists who were used to the theory-based DP-100 may find the shift to engineering a bit jarring. However, one welcome development is that there will be fewer questions about specific Python libraries (like Scikit-Learn or Pandas) and more about architectural workflows.

5. When will AI-300 be generally available?

The AI-300 certification is expected to be available by May 2026. The beta version of AI-300 was launched in March 2026.