PL-900 in 2026: Power Platform Fundamentals and Agents

PL-900 remains Microsoft’s active Power Platform Fundamentals exam in 2026, but the content has changed substantially from older versions of the certification. The current blueprint no longer treats Power BI and Power Virtual Agents as large standalone exam domains. Instead, it emphasizes Power Platform business value, environment management, Power Apps, Power Automate, and agents in Copilot Studio.

That makes older Microsoft PL-900 study material risky when it still presents the product suite as Power Apps, Power Automate, Power BI, and Power Virtual Agents in equal proportions. Power BI remains part of the wider Microsoft analytics ecosystem, but the current PL-900 skills-at-a-glance places much more weight on administration, makers, automation, and agent experiences.

The Microsoft Power Platform Fundamentals credential is still designed as an entry point. It is useful for business professionals, new makers, students, administrators, and technologists who need a clear mental model of how Power Platform creates business solutions without yet specializing in development or architecture.

The current exam starts with business value, not product memorization

The first PL-900 skill area asks candidates to explain the business value of Microsoft Power Platform. That includes understanding how organizations can create applications, automate processes, use Dataverse, build external experiences, and increasingly use AI and agents to support work.

The key word is value. A candidate should be able to look at a business requirement and decide which kind of Power Platform capability fits. A manual approval process may point toward Power Automate. A mobile inspection experience may point toward Power Apps. A structured business-data requirement may point toward Dataverse. A conversational experience with knowledge and bounded actions may point toward Copilot Studio.

This is more durable than memorizing marketing definitions because the platform will continue to evolve. The exam expects candidates to understand how the pieces work together around a business process.

The updated Power Platform architecture is a useful companion because it shows how apps, automation, data, analytics, external experiences, and agents now form one connected solution environment.

Environment management is now a substantial fundamentals domain

PL-900 candidates need more awareness of Power Platform administration than older versions of the exam implied. Environments separate resources, makers, data, and solution lifecycles. Administrators and makers need to understand why development, test, and production should not be treated as one shared workspace.

Environment concepts also connect to governance. Data loss prevention policies can control which connectors may exchange data. Security roles and sharing determine who can access Dataverse and apps. Connection ownership and environment strategy affect whether an automation remains reliable when people change roles.

Fundamentals candidates do not need the depth of a platform administrator, but they should understand why governance exists. Low-code creation can scale quickly, so organizations need boundaries that make experimentation possible without allowing every prototype to become an unmanaged production dependency.

A useful lab is to compare two environments, inspect available resources, and observe how connections, apps, or solutions belong to an environment rather than to Power Platform globally.

Also consider ownership. A solution built by one employee can become business-critical long before the organization formally recognizes it. Environment strategy, managed solutions, connection references, and clear support ownership are what turn low-code speed into sustainable operations rather than a collection of personal assets.

Power Apps is about matching the application style to the work

Power Apps remains a major PL-900 domain. Candidates should understand canvas apps, model-driven apps, common data connections, basic formulas, controls, and the role of Dataverse.

Canvas apps are useful when the user experience needs detailed control over screens and interactions. Model-driven apps begin with Dataverse structure and are a strong fit for record-centric business processes where forms, views, relationships, and security are central.

The important fundamentals skill is not deciding that one style is better. It is recognizing which interaction model fits the business process. A technician collecting field data on a phone may need a very different experience from a service team working through customers, activities, and cases.

AI-assisted creation can accelerate prototypes, but the maker still needs to validate data structure, permissions, formulas, accessibility, and user experience. Generated does not mean production-ready.

Power Automate should be understood as event-driven business logic

Power Automate connects triggers to actions. A record changes, an approval is requested, a form is submitted, a schedule is reached, or another system emits an event. The flow then performs work through connectors, Dataverse, APIs, or other services.

PL-900 candidates should recognize common flow types and know why automation can improve consistency, reduce repetitive work, and connect otherwise separate systems. They should also understand that automation can create operational dependencies.

A flow that sends notifications is relatively low risk. A flow that creates financial records, changes customer data, or invokes another business system has more consequence. Ownership, permissions, error handling, and monitoring become increasingly important as automation moves closer to critical processes.

Build one simple approval or record-driven flow and then deliberately change the trigger or connection. Observing the failure helps explain why administration and automation belong in the same fundamentals exam.

Copilot Studio and agents are now part of core PL-900 knowledge

One of the biggest changes in the current PL-900 blueprint is the explicit weight given to agents in Copilot Studio. Power Virtual Agents has evolved into Copilot Studio, and modern Power Platform solutions increasingly include agents that can answer from knowledge, use tools, call flows, and participate in business processes.

Fundamentals candidates should understand the difference between knowledge and action. An agent that answers from approved information has a different risk profile from an agent that can create a record, start a flow, or call an API.

Agent design therefore introduces questions about identity, permissions, grounding, tools, confirmation, and monitoring even at a fundamentals level. You do not need to become an AI engineer for PL-900, but you should understand why an agent belongs inside the same data and governance model as the rest of Power Platform.

This shift explains why older chatbot material feels incomplete. The current product direction is broader than scripted conversation.

Dataverse remains the shared data and security foundation

Dataverse stores business data in tables with columns, relationships, ownership, and security. It also provides APIs, auditing, solution-aware components, and a common platform that Power Apps, Power Automate, Dynamics applications, and agents can use.

At PL-900 level, candidates should understand why Dataverse is more than a spreadsheet in the cloud. Relationships can enforce a coherent business model. Security roles can control access. Solutions can package components for deployment. The same records can support an app, a flow, and an agent without every component inventing its own data structure.

Connectors extend the model to external services. Standard and custom connectors allow Power Platform to exchange data and invoke capabilities outside Dataverse, but governance policies can limit which services may be combined.

A useful learning exercise is to create a small table, relate it to another table, build a simple app on top, and observe how the same data can participate in automation.

PL-900 is a foundation for specialization, not a substitute for it

Passing PL-900 does not make someone a Power Platform developer, functional consultant, architect, or administrator. The credential establishes vocabulary and architectural awareness so the learner can understand how those roles interact.

The Power Platform certification roadmap helps place the fundamentals credential against more specialized paths. Those paths can change as Microsoft retires and replaces exams, so current official objectives should remain the source of truth.

A learner moving toward professional development should deepen Dataverse, APIs, code, integration, ALM, and security. A functional consultant needs stronger requirements and process design. A maker should become better at app and automation design. Fundamentals is useful because it makes the boundaries visible.

The same principle applies to Applied Skills. A focused lab credential can deepen one technical scenario after PL-900, but it does not automatically replace the broader responsibility represented by a role certification.

Readiness means you can connect a business problem to the right platform capability

A strong PL-900 candidate can take a short business scenario and explain which Power Platform capability is appropriate, where the data belongs, which identity or permission matters, and how the solution would be governed.

Build one miniature end-to-end solution: store simple business data, create a Power App, automate one event, and add a limited Copilot Studio experience that uses approved knowledge or a safe action. The objective is not complexity. It is understanding how the components cooperate.

Use Microsoft role-based certifications when planning what comes next. PL-900 has value when it gives you a correct 2026 mental model of Power Platform: environments and governance around apps, automations, Dataverse, and increasingly agents.

That mental model is more useful than memorizing an outdated four-product list. The platform has evolved, and the fundamentals exam has evolved with it.

A final readiness test is to explain the solution without naming the product first. Describe the business problem, data, users, trigger, actions, and security boundary, then select the Power Platform capability that fits. That sequence demonstrates understanding rather than vocabulary recall, and it remains useful as Microsoft continues changing interfaces.