AWS SAP-C03 Exam Update: Take SAP-C02 Now or Wait?

AWS announced the next AWS Certified Solutions Architect – Professional exam on September 1, 2026. Registration for SAP-C03 opens October 27, the final day to take SAP-C02 is November 16, and SAP-C03 testing begins November 17. AWS says the new exam will launch in every language currently supported for SAP-C02.

This is a version transition, not a reason to discount a credential earned on the current exam. AWS explicitly says candidates who pass SAP-C02 before the deadline keep an active certification through its original expiration date. Your decision should therefore be based on readiness and the skills you want to validate, not on a fear that an SAP-C02 pass will immediately become invalid.

AWS describes the updated role as extending beyond highly available, cost-efficient systems. The confirmed additions include generative and agentic AI, post-quantum cryptography, resilience engineering, cloud-native application patterns, DevSecOps automation, observability, and modern data architectures.

The underlying professional-level decisions still matter: translating complex requirements into architecture, governing multi-account environments, protecting workloads and data, balancing cost against business needs, planning migrations, and designing for continuity. Treat the new topics as additional contexts in which those decisions must be made, not as substitutes for core architecture skills.

AWS has named five domains: cloud-native architecture design and implementation; security, compliance, and governance; cost-optimized architecture design; resilience, migration, and business continuity; and operational excellence and automation. AWS had not yet published the full exam guide when this article was reviewed on September 12. It says the detailed task statements will be available when registration opens on October 27.

That distinction matters. The announcement is enough to begin building relevant skills, but it is not enough to assign precise study percentages or declare that a particular service will dominate the exam. Use the October exam guide as the final scope checklist before converting these themes into a detailed hour-by-hour plan.

AWS identifies 11 new generative and agentic AI skills, including Amazon Bedrock integrations, Amazon Bedrock AgentCore, retrieval-augmented generation architectures, Bedrock Guardrails, and human-oversight workflows. Study these as design choices: what data the system can retrieve, what an agent may do, where authorization is enforced, how unsafe output is constrained, and when a person must approve an action.

The security expansion also names post-quantum cryptography with AWS KMS, including ML-DSA and ML-KEM. At professional level, avoid reducing this to acronym recall. Practice identifying the assets and connections that need protection, the compatibility and migration constraints, and the operational plan for introducing new cryptographic controls without breaking dependent systems.

The resilience additions name AWS Fault Injection Service, AWS Resilience Hub, AWS Application Recovery Controller, and automated runbooks through AWS Systems Manager. A useful lab or design exercise should connect each service to an outcome: test an assumption, assess resilience posture, control recovery or routing, and automate a repeatable operational response.

AWS also calls out pipeline vulnerability scanning, multi-account deployment pipelines, container monitoring, AI and machine-learning metrics, synthetic monitoring, and real-user monitoring. For each scenario, ask what signal proves the workload is healthy, where a control should run, who can act on the result, and how the response works across accounts and environments.

AWS lists serverless data pipelines, durable workflows with AWS Step Functions, service connectivity with Amazon VPC Lattice and Amazon ECS Service Connect, container image security, multi-tenant architecture, and real-time data architecture among the cloud-native additions. It also names lakehouse designs with AWS Lake Formation and Apache Iceberg, real-time and batch pipelines, and cross-account sharing with AWS Clean Rooms.

Do not build a glossary and stop there. Compare failure behavior, scaling boundaries, tenancy isolation, network reachability, ownership, data freshness, governance, and cost. Professional-level practice should make you defend why a pattern meets the whole requirement and why a plausible alternative creates a specific risk or operational burden.

Choose SAP-C02 when you are already near exam readiness, your practice and labs align with the current guide, and you can book with a sensible buffer before November 16. Do not schedule the final available day if an earlier date is realistic; appointment availability, illness, or a needed retake can turn a tight plan into a forced version change.

A useful readiness check is whether you can work through unfamiliar multi-account, migration, security, continuity, and cost scenarios without relying on remembered keywords. If your plan still contains large gaps in those fundamentals, the calendar alone is not a good reason to rush a professional-level attempt.

Waiting makes sense when your realistic test date is after November 16, you are early in your preparation, or you specifically want the exam to validate AI-integrated architecture, modern resilience, cloud-native delivery, and DevSecOps. Registration opens October 27 and general testing begins November 17; AWS lists 75 questions, 180 minutes, and a price of 300 USD for the new exam.

Use the period before the detailed guide appears to strengthen transferable architecture skills and run small labs around the confirmed services. On October 27, compare your notes with the official SAP-C03 task statements, mark retained knowledge and genuine gaps, and revise your plan. That is more reliable than rebuilding your study schedule from announcement headlines.

First, write down the exam version and target date you are actually pursuing. Second, keep a current blueprint checklist beside your notes: SAP-C02 until November 16, or the SAP-C03 guide after AWS publishes it. Third, turn every topic into a design exercise with requirements, constraints, a proposed architecture, failure modes, and verification evidence.

Fourth, review mistakes by the tradeoff you missed—security, resilience, cost, operations, migration risk, data governance, or organizational complexity. Associate-level knowledge is useful groundwork, but professional preparation should require you to reconcile competing requirements across accounts, workloads, teams, and business outcomes.