Imagine you’ve been preparing for the AWS Certified Machine Learning Engineer – Associate (MLA-C01) exam for weeks. You’ve reviewed machine learning concepts, practiced with Amazon SageMaker, and worked through practice exams. Then you hear the news: AWS is updating the certification to MLA-C02.

The AWS Certified Machine Learning Engineer – Associate validates your technical ability to implement machine learning workloads in production and operationalize ML solutions on AWS. In 2026, AWS is updating the exam to better reflect the evolving responsibilities of machine learning engineers, including developments in generative AI, foundation models, and agentic AI.

The most important dates to remember are:

• September 1, 2026 – Registration for the updated MLA-C02 beta exam opens.

• September 28, 2026 – The last day to take the current MLA-C01 exam in English.

So, whether you’re already deep into your MLA-C01 preparation or just starting your AWS machine learning certification journey, understanding what’s changing can help you decide your next move.

In this article, we’ll walk through what’s changing with MLA-C02, the important transition dates, what AWS has confirmed so far, and what these updates mean for your exam preparation.

What Is the AWS Certified Machine Learning Engineer – Associate?

The AWS Certified Machine Learning Engineer – Associate certification validates your technical ability to build, deploy, operationalize, maintain, and monitor machine learning solutions on AWS.

It is designed for professionals working with production ML workloads, including machine learning engineers, MLOps engineers, data engineers, software developers, and data scientists.

The current MLA-C01 exam focuses on four main areas:

• Data Preparation for ML and AI

• ML Model and Foundation Model (FM) Development



• Deployment and Orchestration of ML and AI Workflows

• Operating, Monitoring, and Securing ML and AI Solutions

AWS recommends that candidates have at least one year of experience working with AWS services for machine learning engineering, as well as experience in a related technical role.

AWS Certified Machine Learning Engineer – Associate Is Being Updated

One of the biggest AWS Certification updates in 2026 is the transition from MLA-C01 to MLA-C02.

AWS is updating the certification to reflect how the role of a machine learning engineer continues to evolve. Modern ML engineers are no longer working exclusively with traditional machine learning models. They are increasingly responsible for implementing generative AI applications, foundation models, large language models (LLMs), and agentic AI workflows alongside traditional ML systems.

The updated MLA-C02 exam is designed to reflect these responsibilities while maintaining the core machine learning engineering skills covered by the certification.

Important MLA-C02 Dates

Registration for the updated MLA-C02 beta exam opens on September 1, 2026.

The beta exam will begin delivery on September 28, 2026 and will initially be available in English only.

Meanwhile, candidates who are already preparing for the current exam still have time:

The last day to take the AWS Certified Machine Learning Engineer – Associate MLA-C01 exam in English is September 28, 2026.

The MLA-C01 exam will continue to be available in Japanese, Korean, and Simplified Chinese during the beta period. AWS states that these versions will remain available until MLA-C02 is generally available.

At general availability, the updated exam will be offered in English, Korean, Japanese, and Simplified Chinese.

What’s New in MLA-C02?

The overall domain structure is staying the same, consisting of four core content domains, but AWS is updating the specific skills and technologies covered within those domains.

One of the biggest changes is the increased focus on generative AI and modern AI engineering across both traditional machine learning models and foundation models.

Candidates preparing for MLA-C02 should expect topics involving:

• Generative AI Implementation – ML engineers must build, operationalize, and deploy generative AI solutions, which include selecting Retrieval Augmented Generation (RAG) architecture patterns, configuring vector databases (such as Amazon OpenSearch Service, Amazon RDS with pgvector, and Amazon S3), preparing documents via chunking and metadata extraction, optimizing embedding models, and orchestrating Knowledge Base refresh cycles.

• Foundation Models and Large Language Models – The updated exam expands beyond traditional ML algorithms to include evaluating and selecting pre-trained FMs from Amazon Bedrock, applying customization techniques (fine-tuning, task-specific prompt engineering, continuous pre-training, and model distillation), configuring FM deployment and resource allocation, and conducting model evaluations using NLP metrics (BLEU, ROUGE, BERTScore, semantic similarity), content quality validation, and LLM-as-a-judge frameworks.

• Agentic AI –ML engineers increasingly deploy and configure AI agents capable of multi-step task execution, tool and service integration, state management system implementation, automated agent deployment pipeline orchestration and versioning, and agent coordination monitoring (detecting tool failures, truncated streaming, or coordination errors).

• Amazon Bedrock – Candidates should expect expanded coverage of Amazon Bedrock capabilities—including Bedrock Agents, Knowledge Bases, Guardrails, Prompt Management, Model Evaluation, and AgentCore Observability—working alongside Amazon SageMaker AI (including JumpStart, HyperPod, and Clarify) and other core AWS machine learning services.

• Responsible AI –The updated exam incorporates responsible AI policies, sensitive data protection (PII/PHI masking, redaction, and anonymization), training data validation (prompt-response pair integrity and content safety screening), prompt injection risk mitigation, and automated safeguards against toxicity and hallucinations using Amazon Bedrock Guardrails and SageMaker Clarify.

Despite these additions, AWS has confirmed that no new exam domains are being added. Instead, existing task statements and skills have been updated to reflect current ML engineering and MLOps practices

MLA-C01 vs. MLA-C02

Your decision mainly depends on how far you are into your current preparation.

If you’ve already spent significant time studying for MLA-C01 and feel ready for the exam, you can still take the current English version on or before September 28, 2026.

AWS also confirms that certifications earned through MLA-C01 will remain active through their original expiration date. AWS Certifications are generally valid for three years.

On the other hand, if you’re just beginning your preparation or want your exam to cover newer technologies such as generative AI, Amazon Bedrock, LLMs, RAG, and agentic AI, you may want to consider the MLA-C02 beta exam.

Registration for the beta opens on September 1, 2026.

MLA-C02 Beta Exam Details