AWS Rebuilt Its Certification Program Around AI in 2026 — Here's the New Map

AWS rolled out its biggest certification restructuring in years starting in early 2026, and if you last checked the certification path more than a few months ago, parts of it will look unfamiliar. Here’s what actually changed and how to figure out which certification fits where you are.

The Retirement: Machine Learning – Specialty Is Gone

AWS Certified Machine Learning – Specialty (MLS-C01) retired on March 31, 2026. If you already hold it, your certification stays active for the full 3 years from your original certification date — nothing changes for you personally. But it can no longer be earned or renewed.

Its replacement isn’t a single exam — AWS split the role into a proper career path instead.

The New AI Certification Path

AWS now has four certifications and one specialty-track exam sitting under the AI/ML umbrella, roughly ordered by experience level:

If you’re mapping a study order: AI Practitioner establishes vocabulary and concepts, ML Engineer – Associate builds hands-on implementation skill, and Generative AI Developer – Professional targets production GenAI work specifically. AWS’s exam voucher discounts are designed to chain in that order too — pass AIF-C01, use the discount toward MLA-C01, and so on.

What Else Changed in the Broader AWS Path

  • Generative AI content is now woven into foundational exams, including the Cloud Practitioner exam — you can no longer treat AI as an optional specialty track and ignore it at the entry level.
  • Security – Specialty (SCS-C03) replaced SCS-C02, with the older version’s last day to test being December 1, 2025.
  • AWS launched Lab Maker on Skill Builder — an AI-powered feature that generates instant, personalized hands-on labs from a natural-language description, aimed at closing the gap between reading about a service and actually configuring it.
  • New microcredentials (free, scenario-based, no multiple choice) were added for AWS MLOps and AWS Agentic AI, letting you validate narrow skills without committing to a full certification exam.

Which Certification Should You Actually Target?

  • New to cloud and AI entirely → AWS Certified AI Practitioner. It’s foundational for a reason — don’t skip to a harder exam just because it sounds more impressive.
  • Already scripting, deploying, or supporting ML models on AWS → Machine Learning Engineer – Associate is the direct replacement for what ML Specialty used to validate, just restructured around implementation rather than deep theory.
  • Building GenAI features into production applications → Generative AI Developer – Professional. This is the exam that didn’t really exist in this form before 2026, and it’s where the newest job demand is concentrated.
  • Security-focused and already AWS-certified elsewhere → the refreshed SCS-C03 is worth revisiting even if you passed the old version, given how much of the new content is AI/ML-specific.

Getting Ready

Whichever exam you’re targeting, the underlying AWS services move fast enough that generic prep material goes stale within months. Work from practice questions that reflect the current exam guide, not a version written before the 2026 restructuring.


DailyDebian’s AWS question sets are kept current against the live exam guides — including the new AI Practitioner, ML Engineer – Associate, and Generative AI Developer – Professional exams.

Back to blog