AWS Certified Machine Learning Specialty Retired: What to Take Instead

The AWS Certified Machine Learning Specialty is retired. The last day to take the exam was March 31, 2026, so it’s no longer something you can sit. If you already hold it, AWS says it stays active for 3 years from the date you earned it. The AWS Certified Machine Learning Engineer Associate is the option AWS points to now.

What Happened to the AWS Machine Learning Specialty?

AWS retired the Certified Machine Learning Specialty. On its certification page, AWS says the credential is being retired and gives March 31, 2026 as the last day to take the exam.

This certification had a loyal following. It was a specialty-level exam for people running machine learning and deep learning workloads on AWS, and plenty of job postings named it. Plenty of study guides and courses still advertise it, which is why people keep searching for it.

Today the page works as a reference. It tells you what the credential was, what happens if you hold it and which current AWS exams to look at.

What Did the Exam Look Like?

For the record, here is what AWS published before retirement:

• Questions: 65, multiple choice or multiple response.
• Time: 180 minutes.
• Cost: $300 USD, the final price before retirement.
• Delivery: Pearson VUE testing center or online proctored.
• Recommended experience: 2 or more years developing, architecting and running ML or deep learning workloads in the AWS Cloud.

The exam was longer and more expensive than the associate-level machine learning exam. It also assumed more experience. That’s part of why the new options don’t line up exactly. AWS moved its main machine learning credential down a level, to associate.

I Already Hold the Specialty: What Do I Keep?

You keep it until it expires. AWS says certification holders still have an active certification for 3 years from the date it was earned.

Practical steps:

• Check your expiry date. Find your certification in your AWS certification account and note the date.
• Keep listing it. Your credential is real until it lapses. List the name and the year you earned it.
• Plan ahead. If your employer or clients care about current AWS credentials, choose a replacement before the 3 years run out. Starting early beats a rush later.

Renewal rules for holders are on the AWS certification site, which is the place to confirm them.

What Should You Take Instead?

AWS lists the AWS Certified Machine Learning Engineer Associate as the certification it offers now. Two other current AWS exams sit on either side of it.

AWS Certified AI Practitioner. Foundational, $100, 65 questions, 90 minutes, no prerequisite. Good for newer learners and business roles. Read our AI Practitioner page.

AWS Certified Machine Learning Engineer Associate. The closest match to what the Specialty covered. The current exam, MLA-C01, costs $150 USD, has 65 questions and runs 130 minutes. AWS is switching it to an updated MLA-C02, now in beta. AWS recommends 1 year of ML engineering experience and 1 year on AWS. Our Machine Learning Engineer Associate page has the details and the version change.

AWS Certified Generative AI Developer Professional. $300, 75 questions, 180 minutes. AWS recommends 2 or more years of production experience on AWS or open-source tools, general AI/ML or data engineering experience, and 1 year of generative AI work. See the Generative AI Developer page.

How Do You Choose Between Them?

Match the exam to what you’ve done, not to the Specialty’s old prestige.

• You’ve shipped models on AWS and want to show it: the Machine Learning Engineer Associate is the direct path.
• You build applications on foundation models: the Generative AI Developer Professional matches that work, and its experience bar is close to the Specialty’s.
• You’re newer to AI: the AI Practitioner is the right start. It costs less and asks for no prior experience.
• You’re a long-time ML practitioner with a specialty background: read the associate’s domains before you decide. It may feel like a step down in title, but the content is current.

If a job posting asks for the Specialty by name, the posting may simply be older than the retirement. Ask the recruiter what they’d accept now.

How Do the Current AWS AI Exams Compare?

Here’s a side-by-side using the figures AWS publishes for each exam.

• AI Practitioner: foundational level, $100, 65 questions, 90 minutes.
• Machine Learning Engineer Associate: associate level, $150 for MLA-C01, 65 questions, 130 minutes. The beta of MLA-C02 has 85 questions and 170 minutes.
• Generative AI Developer Professional: professional level, $300, 75 questions, 180 minutes.
• Machine Learning Specialty (retired): specialty level, $300, 65 questions, 180 minutes.

Notice that the Specialty’s time and price match the professional exam, not the associate. If you’re asking which one sits at the highest level now, that’s the Generative AI Developer Professional. It’s also a different subject: it targets generative AI applications, while the Specialty targeted machine learning broadly.

So which exam replaces the Specialty? No single one does. The associate covers the engineering workflow, and the professional exam covers the senior generative AI build work. Which one fits depends on your work.

Does the Retirement Hurt Your Résumé?

Not if you hold the credential. It’s a real certification you earned, and it stays active for 3 years. A hiring manager who sees “AWS Certified Machine Learning Specialty” will read it as a sign of serious AWS machine learning work.

It does mean the credential has a clock on it, and it can’t be extended by retaking the exam. Plan for the date, and add a current credential or a clearly described project on top of it.

If you never held it, the retirement doesn’t count against you. Employers can see that it’s gone, so a current posting is more likely to name the associate or no exam at all.

A few ways to describe your status accurately:

• Holder: list it with the year you earned it.
• Studying for it before it retired: don’t list it. Say you studied AWS machine learning and list the current exam you’re planning.
• Working toward a replacement: name the exam and the date you plan to sit it.

Honest wording helps you. A recruiter can check an AWS credential, and a mismatch costs more than a gap.

What If You Wanted the Specialty for a Career Change?

Many people looked at the Specialty as a way into machine learning work. That was always a hard route, because AWS recommended 2 or more years of hands-on experience before sitting it. It was a credential for people already doing the work.

If that was your plan, the better route now is to build the base first.

1. Learn the foundations. Python, statistics and linear algebra come before any cloud exam.
2. Pick a path. A degree gives you structure, credit and a network. Self-study and projects work if you’re disciplined and your target employers hire that way.
3. Take the AI Practitioner for a first credential, if you want one.
4. Build projects on AWS, and then sit the associate exam once you have the experience AWS recommends.

The ranking list for online machine learning degrees is a place to compare programs if you choose the degree path.

Where Does a Degree Fit?

A degree and an AWS credential aren’t substitutes. A cert doesn’t replace a degree, and a degree doesn’t replace a cert.

An AWS exam checks one platform’s tools and practices. A degree teaches the statistics, programming and modeling behind them, and it gives you college credit. A hiring manager who requires a degree won’t accept an exam in its place. A hiring manager who wants AWS skills won’t assume a degree covers them.

The Specialty’s retirement is a useful reminder of why. Vendor credentials change. Exams get retired, renamed and rebuilt. A degree doesn’t expire, and the theory you learn in one keeps applying across platforms.

If you’re working toward a degree, our ranking lists are the place to start. Online machine learning degrees fit the topic this exam covered most directly. If you want graduate-level depth, see online master’s in artificial intelligence. And if you’re weighing whether to pursue one at all, Is an AI Degree Worth It? lays out the tradeoffs.

What Jobs Did the Specialty Help With, and What Helps Now?

The Specialty pointed toward machine learning engineering, applied ML and data science work on AWS. Those jobs still exist, and the skills still matter even though the exam is gone.

Our machine learning engineer and AI engineer pages cover duties, typical degree levels and pay. Data scientist work also overlaps. For all of them, your projects, your degree and your AWS experience matter more than one credential.

Employers hiring in this area usually look for the same things:

• A degree or equivalent training in a quantitative field.
• Projects that show you can take a model from data to deployment.
• Experience with the cloud platform they use.
• A current credential, if one is easy to point to.

The last item is where the associate exam comes in. It’s the current AWS machine learning credential, so it’s the easiest one to point to.

How Should You Study for the Replacement?

If you studied for the Specialty, you aren’t starting over. Much of what you learned carries into the associate exam, which covers data preparation, model development, deployment and orchestration, and monitoring and security.

A simple plan:

1. Read the current exam guide and compare its tasks with what you already know.
2. Focus on the gaps. Deployment, orchestration and monitoring are where practitioners often have less hands-on time.
3. Read what’s new. AWS says the updated MLA-C02 adds generative AI, agentic AI and foundation models.
4. Practice with exam-style questions for the version you’re booking.

If you never held the Specialty and are looking at it because a course mentioned it, skip it. Study the current exam.

How Do You Check What AWS Offers Today?

AWS changes its lineup, so confirm before you book anything. Open the certification page for the exam you’re considering and look at three things: the exam code, the price and the last-day-to-test notice, if one is shown.

The pages we used for this guide are linked in the sources below. Each one states the cost, the question count and the time limit in AWS’s own words. If a page says an exam is being retired or updated, plan around that date.

It also helps to keep your study materials current. A course written for a retired exam will teach you things the replacement no longer tests, and leave out things it now does.

What Should You Avoid?

Buying a Specialty course. You can’t take the exam, so a course for it won’t help you pass anything.

Assuming your old credential renews itself. AWS says it stays active for 3 years from when you earned it. This page doesn’t cover what happens after.

Treating the associate as a downgrade. It’s a different level aimed at engineering work. For a lot of roles, it’s the credential that matches the job.

Skipping the AWS page. Exams and versions are changing, so confirm the current details before you pay.