AWS Certification in 2026: Current Paths, Exam Transitions and the Skills Behind the Badges

AWS certification in late 2026 spans foundational, associate, professional and specialty credentials, with newer AI-oriented options extending the catalog beyond traditional cloud engineering. The portfolio is broad enough that candidates should not begin by asking which exam is “best.” A better first question is which job they are building toward: cloud practitioner, developer, operations engineer, data engineer, machine-learning engineer, solutions architect, DevOps engineer, security specialist or a business professional responsible for AI adoption.

The foundational layer includes AWS Certified Cloud Practitioner and AWS Certified AI Practitioner. These credentials provide broad context, but AWS does not require them before an Associate exam. Candidates with meaningful hands-on experience can move directly to the role-aligned certification that matches their work.

Associate certifications separate architecture, development, operations, data and ML

AWS Certified Solutions Architect – Associate emphasizes design choices across compute, networking, storage, security, resilience and cost. Strong preparation compares alternatives instead of memorizing service summaries: managed versus self-managed, synchronous versus asynchronous, single-AZ versus multi-AZ, public versus private connectivity, relational versus key-value storage, and performance versus cost.

AWS Certified Developer – Associate focuses more heavily on building, testing, deploying and debugging cloud applications. Candidates should connect Lambda, APIs, queues, databases, identity, configuration and monitoring into complete application flows. The upcoming DVA-C03 update reinforces the same systems view while adding current development practices, including AI-assisted development and security for AI-service interactions.

Operations has also changed branding. AWS Certified CloudOps Engineer – Associate replaced the former SysOps Administrator Associate identity. The role still centers on deployment, monitoring, reliability, security and operational response, so candidates need to understand what normal looks like, how failure appears in telemetry and how to restore service safely.

AWS Certified Data Engineer – Associate focuses on ingestion, transformation, storage, data operations and quality. AWS Certified Machine Learning Engineer – Associate moves into implementing and operationalizing ML workloads. These paths overlap in pipelines and operations, but the ML credential adds model-oriented concerns such as training, deployment, monitoring and the lifecycle of production inference.

The September 2026 exam transitions need exact date awareness

AWS published a major certification update on September 1, 2026. For Machine Learning Engineer – Associate, beta registration for MLA-C02 is already open. The last day to take MLA-C01 in English is September 28, 2026, while Japanese, Korean and Simplified Chinese delivery of MLA-C01 continues during the beta period until MLA-C02 general availability on January 14, 2027. Candidates studying at this transition point should choose the exam version deliberately rather than mixing objectives.

Solutions Architect – Professional is also changing. AWS says registration for SAP-C03 opens October 27, 2026, the last day to take SAP-C02 is November 16, and SAP-C03 general availability begins November 17. The updated blueprint expands the modern architecture context with areas such as AI/ML integration, current security practices, resilience and automation while retaining the core requirement to design optimized AWS solutions.

Developer – Associate follows shortly afterward. DVA-C03 registration opens October 27, the last day for DVA-C02 is November 30, and DVA-C03 general availability begins December 1, 2026. Candidates already prepared for DVA-C02 have a clear completion window; candidates at the beginning of a longer plan should compare the updated objectives before deciding which version to target.

Professional credentials expect decisions across entire systems

AWS Certified Solutions Architect – Professional expects candidates to design across organizations, accounts, networks, migrations, governance and complex workloads. At this level, an answer is rarely correct because it uses the most advanced service. It is correct because the design satisfies the stated constraints with an acceptable balance of reliability, security, performance, operational effort and cost.

AWS Certified DevOps Engineer – Professional is the advanced path for software delivery, infrastructure automation, monitoring, incident response and operational governance. Candidates should understand delivery as a controlled system: source changes move through testing, artifacts, deployment, verification and rollback, while infrastructure and observability are managed with the same discipline.

AWS Certified Generative AI Developer – Professional represents the newer production-AI branch of the catalog. It is most useful to developers who already understand application architecture and now need to integrate foundation models, retrieval, agents, security, evaluation and operations into systems that can be governed in production.

A practical way to prepare for professional exams is to review architectures against a consistent set of architecture trade-offs. Ask what fails, what becomes expensive, what is difficult to operate, where trust is established and how the design changes when scale or regulatory requirements change.

Security and networking remain deep specialties, but one is retiring

AWS Certified Security – Specialty continues to validate advanced security capability. The current SCS-C03 path should be studied through identity, detection, incident response, infrastructure security, data protection and governance. It is easier after candidates can already reason about accounts, roles, policies, network boundaries, encryption and workload architecture.

Security candidates benefit from following a coherent AWS security learning path rather than jumping directly into specialty terminology. The same principle applies to networking, but the lifecycle is different: AWS Certified Advanced Networking – Specialty is being retired, with December 31, 2026 as the last day to take the exam. Credentials earned before retirement remain active for their normal validity period.

The retirement does not make advanced networking knowledge less important. Hybrid connectivity, routing, DNS, security controls, multi-region connectivity and network troubleshooting continue to appear in architecture and operations roles. Candidates who cannot reasonably finish the specialty exam before retirement should shift the goal from collecting the retiring badge to preserving the underlying networking capability.

Identity, networking and observability connect nearly every technical exam

Identity is one of the most reusable AWS skills. Candidates should understand IAM policies, roles and least privilege as design mechanisms. The important questions are who or what needs access, which actions are required, which resources should be reachable, how temporary credentials are obtained and how excessive permission is prevented.

Networking is similarly cross-cutting. VPCs, subnets and routing affect application reachability, security, hybrid connectivity, service endpoints and failure domains. Candidates should be able to trace a connection end to end and identify where routing, name resolution, security policy or service configuration could stop it.

Operational visibility completes the picture. Monitoring, logs, metrics and alarms are not optional extras added after deployment. They provide the evidence needed to detect failure, measure user impact and decide whether a change improved the system. CloudOps, Developer, DevOps and architecture candidates all benefit from being able to turn telemetry into an operational diagnosis.

Automation and event-driven design should be practiced as end-to-end flows

Infrastructure as code, deployment pipelines and repeatable configuration reduce drift and make cloud environments auditable. The concepts behind infrastructure as code matter because they shift infrastructure from a sequence of manual actions to a versioned desired state. Candidates should understand modules, state, change previews, drift and the consequences of modifying shared infrastructure.

For developers, event-driven architecture recurs across Lambda, API Gateway, queues, topics, event buses and workflow services. The key design questions concern coupling, retries, ordering, duplicate processing, failure destinations and observability. A queue is not merely a service to memorize; it is a way to change the failure and scaling relationship between producers and consumers.

Build small systems that expose these behaviors. Deploy an application from code, intentionally introduce a bad change, observe how the pipeline and telemetry respond, then roll it back. Create an asynchronous workflow, make the consumer fail and inspect retry behavior. Certification questions become easier when the candidate has seen the architecture behave under stress.

Machine learning is moving from a retired specialty to an active associate path

The former AWS Certified Machine Learning – Specialty exam retired on March 31, 2026. Existing credentials remain active for their normal three-year period, but new candidates can no longer use that exam as the entry point. AWS now directs technical ML certification toward Machine Learning Engineer – Associate.

That shift reflects a broader change in how cloud ML work is evaluated. Candidates need to understand data preparation, model development, deployment and operations as one lifecycle. A model that performs well in a notebook can still fail in production because data changes, latency is unacceptable, access controls are weak or monitoring cannot detect degradation.

The current MLA transition therefore deserves careful planning. Someone already prepared for MLA-C01 in English has only the late-September window remaining; a new learner will generally be better served by studying against MLA-C02. Version-specific details matter, but the durable skill is the ability to operationalize ML workloads with reliable data, deployment and monitoring practices.

Build one role-aligned path instead of collecting unrelated badges

A coherent AWS path compounds skills. A solutions architect can move from Associate architecture into Professional architecture and security; an operations engineer can progress from CloudOps toward DevOps; a developer can deepen application delivery and generative AI; a data professional can combine Data Engineer with machine learning; and business leaders now have AI-oriented options without pretending to be technical implementers.

Labs should mirror the target role. Architects should compare designs and failure modes. Developers should build and debug applications. Operations candidates should detect and recover from incidents. Security candidates should implement and investigate controls. Data and ML candidates should operate pipelines and workloads over time. This produces evidence that the candidate understands the system, not merely the exam vocabulary.

Finally, lifecycle verification should be part of every AWS study plan in 2026. Check the exact exam code, registration date, last testing date and current guide before building a months-long schedule. The catalog is changing quickly, but the underlying cloud concepts repeat. Mastering those relationships is more valuable than accumulating disconnected badges whose names may change before the skills do.

100% Real & Latest Amazon Certification Practice Test Questions and Exam Dumps will help you prepare for your next exam easily. With the complete library of Amazon Certification VCE Exam Dumps, Study Guides, Video Training Courses, you can be sure that you get the latest Amazon Exam Dumps which are updated quickly to make sure you see the exact same questions in your exam.