A lot of the AWS study material sitting on shelves right now points people at exams that changed or vanished, and most of them have no idea yet. AWS spent 2026 quietly rebuilding its certification lineup around one thing, and that thing is AI. One flagship exam retired. A new professional credential showed up. The security exam got torn down and put back together, and even the entry-level test, the one total beginners take first, now asks about generative AI. If your AWS plan is more than a few months old, give it a second look before you pour another weekend into it.
None of this is a reason to panic. AWS still runs about a third of the cloud market, and its certs are still the ones employers ask for by name. If anything the new lineup is sharper than the one it replaced. You just have to know which exams are live and which ones are ghosts. Here’s what moved, and what to do about it.
The Machine Learning Specialty is dead as of March 31, 2026. AWS replaced it with a role-based AI track that runs from the AI Practitioner up to the brand-new Generative AI Developer Professional. If that old exam was the plan, the plan is over.
What Changed on the AWS Certification Map in 2026
AWS moved several pieces at once, so it helps to see them on one board before reading the detail. The table below covers what changed, where it stands now, and what you should do about it if it touches your plan.
The Machine Learning Specialty Is Gone. Here Is What Replaced It.
This is the change that matters most, because it rewires the entire AI path. The AWS Certified Machine Learning Specialty held its last exam on March 31, 2026. If you already earned it, you keep it for the full three years from your test date, so nothing you have is lost. But it’s not a path for new candidates anymore, and there’s no sadder waste of study time than grinding through an exam you can no longer schedule.
The reason behind the retirement tells you exactly where AWS is heading. That old Specialty was built for a world where teams trained models from scratch, deep in algorithms and training pipelines. Almost nobody works that way now. Teams reach for managed foundation models through services like Amazon Bedrock and wire AI features into their apps instead of owning the whole model lifecycle. So AWS scrapped one big exam and built a ladder that matches how AI teams actually hire.
The New Generative AI Developer Professional
The headline addition is the AWS Certified Generative AI Developer Professional, and it came out of beta in 2026 as a full standard exam. It’s built to prove you can take a foundation model and turn it into something a business actually ships, working with RAG architectures, vector databases, and Bedrock integration patterns. Three years ago this job barely existed. Now AWS has a professional credential pointed right at it.
And the demand isn’t hype. There are nowhere near enough engineers who can actually build and run generative AI in production rather than talk about it in meetings, and that gap shows up fast in what these people get paid. If you write code, you’ve got AWS time behind you, and you want a credential aimed straight at where the money is moving, this is the one to keep an eye on. The professional-tier salary patterns on AWS credentials give you a feel for what that kind of scarcity does to a paycheck.
SysOps Became CloudOps, and Security Got an AI Refresh
Two more changes matter if either one sits on your path. First, the operations route got a new name and a wider job description. AWS replaced the old SysOps Administrator Associate with the AWS Certified CloudOps Engineer Associate, which fits how much broader operations work has gotten since the SysOps title was written. You’ll still hear people say take SysOps because it has the labs. That advice is years out of date. CloudOps Engineer Associate is the operations credential now, so plan around that one.
Second, the Security Specialty got rebuilt. The current version, SCS-C03, leans hard into emerging tech, with a dedicated focus on generative AI and machine learning security. This one isn’t going anywhere, and honestly it matters more than it did before, because securing AI workloads is now part of the day job rather than a future problem. Study the SCS-C03 objectives and skip anything older. If you want a feel for what hands-on AWS security actually looks like, our fast AWS security wins are a good place to start.
Even the Cloud Practitioner Now Tests AI
If you want the clearest signal of where all this is going, look at the bottom of the ladder, not the top. The Cloud Practitioner is the exam thousands of beginners take as their very first cloud credential, and AWS just refreshed its guide to fold in generative AI concepts. Even the most basic test in the catalog now expects you to know some AI fundamentals on top of the usual services and billing.
That’s the whole strategy in one move. AWS is treating AI literacy as the price of admission, not as a fancy specialty you bolt on once you’re senior. When even the beginner exam asks about generative AI, the company is telling you that cloud and AI stopped being separate subjects. They’re one subject now.
What This Means for Your AWS Certification Plan
So what do you actually do with all this? A few things. Cross the Machine Learning Specialty off your list, because you can’t take it, and pick up the role-based AI ladder instead. Then, whatever your real lane is, architect, developer, operations, plan to add some AI literacy on top of it, since AWS has now baked AI into everything from the first exam upward. The AI Practitioner is the cheap, fast way to get that on your record, and it works no matter what you do for a living. Build your base with a core credential, which for most people is the Solutions Architect Associate, then branch toward the AI ladder or a Specialty that fits the job you want. One more wrinkle. If you already hold AWS and you are eyeing a second cloud, that is a different decision entirely, and we walk through it in do I need Azure if I already have AWS. The catalog moved, but the strategy did not. Pick the cert that matches the job you’re chasing, and make sure it’s one AWS still sells.
One practical note on money and timing. AWS exams still run $100 at the foundational level, $150 at associate, and $300 for professional and specialty, every credential stays valid for three years, and passing one exam earns a fifty percent discount voucher toward the next. That voucher chaining matters more now, because the role-based AI ladder means most people will stack two or three exams rather than one, and the discounts add up across the climb.
Frequently Asked Questions
Is the AWS Machine Learning Specialty still available?
No. The AWS Certified Machine Learning Specialty held its last exam on March 31, 2026. If you earned it before then, it stays valid for three years from your test date. New candidates should use the role-based AI path instead, starting with the AI Practitioner and moving to the Machine Learning Engineer Associate.
What is the AWS Generative AI Developer Professional certification?
It’s a professional-level AWS credential, launched out of beta in 2026, that validates a developer’s ability to build production generative AI applications. The exam covers foundation model integration, RAG architectures, vector databases, and Amazon Bedrock, and it targets developers and AI engineers shipping real AI solutions rather than experimenting.
Did AWS rename the SysOps Administrator certification?
Yes. The AWS Certified SysOps Administrator Associate has been replaced by the AWS Certified CloudOps Engineer Associate, reflecting a broader cloud operations role. New candidates on the operations path should plan around CloudOps Engineer Associate rather than the retired SysOps exam.
Do I need the AWS AI Practitioner certification?
It’s increasingly worth holding regardless of your role. AWS has added generative AI content across its catalog, down to the Cloud Practitioner, so AI literacy is becoming a baseline expectation. The AI Practitioner is a foundational, lower-cost way to validate that literacy whether you architect, develop, or run operations.
Are my existing AWS certifications still valid after these changes?
Yes. Any AWS certification you have already earned stays valid for its full three-year term, including ones tied to retired exams like the Machine Learning Specialty. A retirement only stops new candidates from taking that exam version; it does not affect credentials already earned.
CEO | Training Camp
Christopher D. Porter is a dynamic marketing executive and visionary leader, celebrated as an early adopter of internet technologies for innovative lead generation strategies. Continuing his career as the CEO of one of the leading IT and Cybersecurity Certification Training companies, he has consistently harnessed digital innovation to drive business growth and market transformation.