CertifyDeck
DP-100Retired exam

DP-100 retired: the current alternative (AI-300) and what to keep

DP-100 and Azure Data Scientist Associate retired on 1 June 2026. What happens to your certification, what AI-300 measures and which part of your study still applies.

3 min readReviewed against the official syllabus
DP-100 retired: the current alternative (AI-300) and what to keep

Microsoft retired exam DP-100 and the Azure Data Scientist Associate certification on 1 June 2026, so it can no longer be booked or passed. The successor in this area is Machine Learning Operations Engineer Associate, with exam AI-300, more focused on operationalizing models and generative AI. Here is what to keep from your preparation and what to add.

1

What happens to DP-100 and to your certification?

The content on Azure Machine Learning, data preparation, training, pipelines and deployment is still instructive, but it no longer leads to an available exam. If you earned the credential before the retirement, Microsoft keeps its record under its policy for retired certifications; what you cannot do is renew it.

2

The new direction: MLOps and AI-300

Microsoft has positioned Machine Learning Operations Engineer Associate as the successor, with a stronger focus on operationalizing machine learning and generative AI solutions. The AI-300 outline covers MLOps infrastructure (15–20%), the machine learning model lifecycle and operations (25–30%), GenAIOps infrastructure (20–25%), quality assurance and observability for generative AI (10–15%) and optimizing generative AI systems (10–15%).

Do not assume all of DP-100 carries over one to one. Check the current AI-300 guide and measure your gaps before reusing an old calendar; the AI-300 outline by domain goes through it block by block.

3

What knowledge you can keep

Keep Python, MLflow, experimentation, model registry, endpoints, data control and evaluation. Add CI/CD automation, reproducible infrastructure, quality monitoring, security and operating generative workloads.

A useful project: train or integrate a model, version it, deploy it with controls and detect its degradation. That path shows competence even when the exam code changes.

4

What you can practise today on CertifyDeck

DP-100 is no longer in the catalog: we do not offer preparation for an exam that cannot be booked. Its successor is: the AI-300 page has practice by domain with an explanation on every question. If your interest is applied AI on Azure rather than operations, look at AI-103, Azure AI Apps and Agents Developer.

Frequently asked questions

When did DP-100 retire?

On 1 June 2026, together with the Azure Data Scientist Associate certification.

Which exam replaces DP-100?

For this area, Microsoft points to AI-300 (Machine Learning Operations Engineer Associate), centred on operationalizing models and generative AI.

Is what I studied for DP-100 still useful?

Yes: Python, MLflow, training, registering and deploying models are the base of AI-300. What changes is the emphasis on CI/CD, monitoring and lifecycle.