AWS Certified Machine Learning – Specialty to Retire March 31, 2026 as AWS Replaces It with Three Role‑Focused AI/ML Certifications
AWS Certified Machine Learning – Specialty will retire on March 31, 2026; AWS replaces the single specialty with AI Practitioner, Machine Learning Engineer – Associate, and Generative AI Developer – Professional.
What AWS announced and why the change matters
The AWS Certified Machine Learning – Specialty certification—long a single, catch‑all credential for cloud professionals working with AI and machine learning on AWS—will no longer be available after March 31, 2026. The move is part of a broader restructuring of AWS’s AI and ML certification portfolio that replaces the one specialty with multiple, role‑focused credentials designed to reflect how teams and job roles actually work with AI in production. If you already hold the Machine Learning – Specialty credential, your certification remains valid until its original expiration date, but new candidates must pick from the updated pathways. (aws.amazon.com)
This shift matters because it signals a change in how cloud providers and the market define AI competence: away from a single, monolithic designation and toward a set of discrete skills aligned with distinct responsibilities—foundation knowledge, model engineering, and generative AI application development. For organizations hiring talent or planning learning investments, the new map aims to make role matching between certification and job expectations more explicit.
The new AWS AI/ML certification pathways
AWS has announced a multi‑tiered certification structure to take the place of the Machine Learning – Specialty, aligning each credential with clearer scopes and responsibilities:
- AWS Certified AI Practitioner (Foundational): Focused on core AI/ML concepts, basic AWS services used for AI, and non‑technical understanding that supports informed conversations about AI within organizations.
- AWS Certified Machine Learning Engineer – Associate (Associate): Oriented toward practitioners who build, train, validate, and deploy ML models—emphasizing model lifecycle, feature engineering, and production concerns.
- AWS Certified Generative AI Developer – Professional (Professional): Targets developers building applications that use generative models—prompt engineering, retrieval‑augmented generation (RAG), model selection and orchestration, and service integration.
AWS also positions an AWS Certified Data Engineer – Associate as a complementary credential for those focused on data pipelines, transformation, and operational data engineering that underpins ML workloads. (aws.amazon.com)
Collectively, this set reflects common role boundaries: business and foundational literacy, hands‑on ML engineering, and the specialized demands of generative AI systems.
Important dates and what they mean for candidates
The final day to take the AWS Certified Machine Learning – Specialty exam is March 31, 2026; after that date the exam will be removed from the catalog and new candidates cannot earn that certification. Current holders keep the credential until the original expiration date printed on their AWS Certification account. AWS has published exam guides and prep plans for the replacement credentials and is staging exam launches and registration windows across late 2025 and 2026 for updated or renamed exams. (aws.amazon.com)
If you’re on a certification timeline, the practical takeaway is straightforward: schedule any final attempts at MLS‑C01 before the retirement date, or pivot your study plan toward the new credentials now.
How to decide which new path to take
Choosing among the new AWS AI/ML credentials should be driven by your day‑to‑day responsibilities and career goals.
- If your role is non‑technical or managerial and you need to speak credibly about AI project feasibility, governance, or vendor choices, the AI Practitioner credential is the most relevant.
- If you design, train, validate, and operate machine learning models or manage MLOps pipelines, the Machine Learning Engineer – Associate aligns with those skills.
- If you build applications that integrate large models, design prompts and retrieval systems, or implement generative AI features in products, the Generative AI Developer – Professional is the appropriate advanced path.
- If your work centers on ETL, streaming, or preparing data for ML, consider the Data Engineer – Associate to demonstrate end‑to‑end pipeline competence.
For candidates who had already been preparing for the legacy Machine Learning – Specialty, the ML Engineer – Associate is the closest successor in content and intent; however, expect updated coverage for modern services and generative workflows that reflect production realities. (aws.amazon.com)
If you already hold Machine Learning – Specialty
Current MLS‑C01 holders do not lose their credential immediately; existing certifications remain valid until their scheduled expiration. When your renewal window arrives—or if you want to expand your skills—you’ll need to map your existing credential onto one of the new certifications. The sensible choices are ML Engineer – Associate for engineers focused on model lifecycle and operations, or Generative AI Developer – Professional for those whose work now centers on generative model applications. AWS’s guidance encourages credential holders to continue along the expanded portfolio to keep skills current. (aws.amazon.com)
If you were studying for MLS‑C01: practical exam prep choices
For candidates deep into MLS‑C01 preparation with an exam date approaching the retirement deadline, assess your readiness realistically. If you can schedule and pass before March 31, 2026, taking the existing exam makes sense. If not, shifting study time to the ML Engineer – Associate pathway—while incorporating newer generative AI topics where relevant—will better align your knowledge with the updated exam landscape.
Study strategy tips:
- Focus on hands‑on labs and production scenarios: AWS emphasizes practical skills such as model deployment, inference optimization, and operational monitoring.
- Practice with modern services: include managed model services, Bedrock‑style generative stacks where applicable, and RAG architectures in your mock projects.
- Validate readiness with practice exams and an 80%+ target on domain‑aligned mock tests before booking an associate/professional exam. (aws.amazon.com)
Related certification changes you should track
The AI/ML recertification is part of a wider set of certification updates at AWS and across other cloud providers:
- SysOps Administrator → CloudOps Engineer (SOA‑C03): AWS has rebranded and updated the SysOps Administrator – Associate into the CloudOps Engineer – Associate with new domain coverage that reflects modern operational responsibilities such as container visibility, automation, and IaC. Registration windows and retirement dates for the older exam were published as part of AWS’s exam roadmap. (aws.amazon.com)
- Security Specialty update (SCS‑C03): The AWS Certified Security – Specialty exam has been revised to expand coverage of AI and ML security topics, including generative model risks and detection/incident response considerations. Candidates should expect new domains that address model governance and threat detection related to AI workloads. (aws.amazon.com)
- Microsoft certification changes: Microsoft is also migrating traditional security and developer certifications toward AI‑focused replacements. For example, AZ‑500 (Azure Security Engineer Associate) is scheduled for retirement with a Cloud and AI Security Engineer (SC‑500) replacement entering beta and production windows in 2026—evidence that major vendors are aligning credentials to AI responsibilities. (techcommunity.microsoft.com)
For practitioners who manage cross‑cloud hiring, training, or role definitions, these synchronous shifts mean you should review competency frameworks and job descriptions to emphasize AI lifecycle, model security, and MLOps fluency.
What the new exams test and how they differ from the legacy specialty
Where the legacy Machine Learning – Specialty tried to cover the full AI/ML stack under one banner, the new design separates responsibilities and depth:
- Breadth versus depth: the AI Practitioner offers breadth and business‑level fluency; the ML Engineer – Associate examines day‑to‑day engineering tasks and lifecycle management; the Generative AI Developer – Professional digs into advanced design patterns for prompt‑centric and retrieval‑augmented systems.
- Modern service coverage: exam guides for the new credentials explicitly include newer managed services and generative AI tooling, reducing the risk of exam content lagging behind product availability.
- Role alignment: each exam has domains tied to real job tasks—data preparation and feature engineering for engineers, prompt design and model orchestration for generative developers, and governance and risk understanding for practitioners. (aws.amazon.com)
This modular approach should make it easier for employers to interpret certificates as proxies for relevant on‑the‑job skills.
What this means for hiring, training, and team design
The restructured certification portfolio changes how employers can build competency ladders:
- Recruitment and job postings can list specific AWS credentials that match required responsibilities, for example “AWS Certified Machine Learning Engineer – Associate preferred” for ML platform engineers.
- Training budgets can be more targeted: buying role‑specific courseware, simulations, and hands‑on practice for the skills that matter.
- Teams can design career ladders that map from AI Practitioner (for business and product roles) into ML Engineer tracks or generative AI specializations, reducing the need for a one‑size‑fits‑all credential.
For organizations, the risk of mismatched expectations should decrease—certifications will better predict candidate suitability for operational or development roles on AI projects.
Implications for developers and platform teams
For developers and platform engineers, the most immediate implication is the need to broaden practical skills around production ML and generative systems:
- MLOps practices (CI/CD for models, feature stores, automated retraining schedules) will be emphasized in ML Engineer pathways.
- Observability and incident response for model behavior—including model drift detection, data quality alerts, and retraining triggers—are now testable skills.
- Generative AI development requires knowledge of prompt engineering, context retrieval, hallucination mitigation, and cost/performance tradeoffs when integrating large models with application logic.
These are not purely academic topics—companies are asking teams to deliver production‑grade behavior, explainability, and secure model operations, and certifications will now test those competencies more explicitly. This makes practical, project‑based experience and portfolio work more valuable than theory alone. (aws.amazon.com)
How to plan a practical learning path right now
If you’re mapping a learning calendar, consider the following approach:
- Define your role goals: practitioner, ML engineer, or generative developer.
- Start with fundamentals: foundational AI/ML concepts, cloud native data handling, and statistics—AI Practitioner covers many of these.
- Move to hands‑on engineering: build experiments that train models, deploy endpoints, and automate retraining; add infrastructure as code and CI/CD pipelines for models for ML Engineer prep.
- Add generative AI projects: implement a RAG pipeline, integrate a managed model API, and measure hallucination and latency tradeoffs for Generative AI Developer readiness.
- Use official exam guides and AWS Skill Builder practice resources, and target practical labs that can serve as demonstrable portfolio pieces. (aws.amazon.com)
Allocating time to real projects—rather than memorization—will pay dividends both for passing exams and for real‑world capability.
Broader industry context and strategic implications
AWS’s recasting of certification taxonomy is part of a larger industry trend in 2025–2026: vendors are re‑scoping credentials to reflect AI‑first job roles, embedding security and governance into clouds’ credentialing, and elevating generative AI as a distinct engineering discipline. Microsoft’s parallel updates on Azure certifications underscore that this is not an AWS‑only phenomenon. For vendors, the benefits are clear: better role alignment, clearer training pathways for enterprise customers, and certifications that more closely match modern job tasks.
For businesses, the change shifts the emphasis from hiring for a label to hiring for demonstrable project experience. HR and L&D teams should update skills matrices, and engineering managers should prioritize internal training that maps to updated exam domains. For vendors of training content, there is an immediate opportunity—and obligation—to refresh curricula to include MLOps patterns, model security controls, and generative‑AI pipelines. (aws.amazon.com)
Practical next steps for candidates and teams
- If you planned to sit MLS‑C01, decide now whether to take it before March 31, 2026, or redirect effort to the successor exams.
- Review AWS’s exam guides and the official prep plans on AWS Skill Builder to understand new domain coverage and required hands‑on competencies.
- For hiring and upskilling, map existing job descriptions to the new certifications and incorporate project‑based assessments that mirror production tasks. (aws.amazon.com)
The real value of certification is the applied capability it represents; employers should test for those abilities in technical interviews and take‑home assignments rather than relying solely on titles.
AWS’s certification redesign signals a practical reorientation of credentialing toward role specificity and modern AI production needs. The retirement of AWS Certified Machine Learning – Specialty closes a chapter in how cloud AI expertise has been recognized, but the new path is explicitly intended to bring certification closer to the day‑to‑day realities of AI development, deployment, and security. Organizations, trainers, and practitioners should use the retirement deadline to review roadmaps and ensure that learning plans and hiring criteria reflect role‑specific competencies being tested in 2026 and beyond. (aws.amazon.com)
Looking ahead, expect certification roadmaps to continue evolving as AI tooling, model governance expectations, and regulatory pressures change: skills around responsible AI, model provenance, explainability, and cross‑cloud integration are likely candidates for future emphasis, and vendors will keep iterating on exam coverage to keep pace with the platform features practitioners use every day.
