top of page

Microsoft Solutions Partner   |  15+ years of CRM experience  |  Trusted by 1000+ D365  crm users

Logo for expert ASG Dynamics 365 CRM support services.

Dynamics 365 2026 Release Wave 2: What Actually Changes for Your Team

Maria Delia Proteasa
Sep 15
4 min read

September 15, 2026


In this article:



Every release wave, Microsoft attempts something incredibly difficult:


Explain hundreds of product changes, across dozens of industries, to thousands of different organizations.


The result is usually accurate.

But not always useful.


A Field Service dispatcher in Rotterdam doesn't need to know everything. A Sales Operations manager in Toronto doesn't either.


They need to know:


  • What changed?

  • What affects their team?

  • What should they test before rollout?


That's the gap this article exists to close.


Dynamics 365 2026 Release Wave 2 promotional graphic featuring AI agents, governance, analytics, Sales, Customer Service, Field Service, and Power Platform workloads, highlighting autonomous Copilot capabilities and stronger administrative controls.
Dynamics 365 Release Wave 2 is not just about new features. It's about deciding where AI should act, where people should stay in control, and how governance keeps pace with automation.

Why This Release Matters Now


The official Dynamics 365 2026 Release Wave 2 documentation is arriving later than usual.


Meanwhile, many organizations are still absorbing features introduced during Release Wave 1.


That doesn't mean you should wait.


Several capabilities reaching general availability right now deserve attention before

Wave 2 arrives on top of them.


We'll update this guide when Microsoft's full Wave 2 documentation is published.


For now, these are the changes worth paying attention to.



What Changed in Dynamics 365 2026 Release Wave 2?


A clear pattern is emerging across Dynamics 365.


Copilot is no longer focused only on summarizing information.


It is starting to take actions.


That's a significant shift.


Instead of helping users work faster, Dynamics is increasingly allowing AI agents to perform parts of the work themselves.


Let's break it down by workload.



Dynamics 365 Sales


  • Natural-Language Search is Now Generally Available


Sales users can search, filter, and sort records using plain language instead of creating complex views.

Why it matters: Reps spend less time building views and more time working opportunities.


  • Form Fill Assist and Smart Paste


Dynamics can now automatically populate and clean information on lead and opportunity records.

Why it matters: The quality and consistency of CRM data will increasingly depend on AI-assisted entry rather than purely manual updates.


  • Sales Skills Inside Microsoft 365 Copilot


Dynamics 365 Sales skills are now available inside Microsoft 365 Copilot Cowork.

Microsoft has confirmed this is enabled automatically for eligible users.

Why it matters:This may affect how sales teams access CRM data without ever opening Dynamics directly.



Dynamics 365 Customer Service


  • Sentiment Indicators on Case Forms


    Customer sentiment is now surfaced directly alongside case information.

    Why it matters: Agents gain additional context before responding, but teams should understand how these indicators are generated before relying on them.


  • Quality Evaluation Agent


    Microsoft is introducing AI-powered evaluation of agent communications.

    Why it matters: Organizations may find AI influencing quality assessments before they've updated their own evaluation standards.


  • AI Credit Monitoring

    Supervisors now receive tools to estimate AI credit usage.

    Why it matters: Copilot is becoming a managed resource that needs capacity planning just like storage or licensing.



Dynamics 365 Field Service


Bulk Booking Management


Dispatchers can reassign and move multiple bookings simultaneously.

Why it matters: A seemingly small update that can significantly improve daily scheduling efficiency.


Scheduling Operations Agent


Microsoft is introducing an agent that automates optimization decisions.

Additional multi-resource automation is expected later in the roadmap.

Why it matters: This is another example of AI moving from recommendation to execution.


Mobile Experience Improvements


Field Service mobile performance enhancements continue rolling out.

Why it matters: Users may experience interface changes without administrators making any configuration changes.



Power Platform Administration Updates


New Usage Analytics Page


Administrators gain improved visibility into adoption and capacity metrics.

Practical question: Who's actually reviewing this data inside your organization?

If the answer is nobody, the feature provides little value.


Expanded Agent Security Controls


Microsoft is introducing stronger controls for AI agent management.

Why it matters: As agents become more autonomous, governance becomes more important.


Guest Access Controls


Administrators can now explicitly allow or block guest access.

Why it matters: This closes a long-standing governance gap that previously required workarounds.



The Bigger Story: Copilot Is Becoming More Autonomous


This is arguably the most important takeaway of the entire release cycle.


Across Sales, Customer Service, and Field Service:


  • Agents create records

  • Agents evaluate performance

  • Agents propose scheduling decisions

  • Agents take actions automatically


The conversation is no longer:

"Can Copilot summarize this?"


It's becoming:

"Should Copilot do this?"


That's a fundamentally different governance discussion.



Why These Updates Matter in Real Environments


Historically, the biggest problems after an update were easy to miss.


A hidden field.


A deprecated API.


A security permission.


This wave introduces a different challenge.


Several new features actively perform work that people used to perform manually.

That isn't necessarily a problem.


But it does mean organizations should understand:


  • What actions agents can take

  • What data agents can access

  • What assumptions agents are operating under


Waiting until production users discover those answers is rarely ideal.



What to Check Before Release Wave 2 Arrives


Before rollout reaches your tenant, review:


✅ Which Copilot features are enabled automatically

✅ AI credit allocation and consumption

✅ Workflows that agents may now influence

✅ Security roles and sensitivity labels

✅ Guest and external access settings


These checks can prevent surprises later.



Sandbox Testing Checklist


Before enabling new AI-driven capabilities broadly:

Test Agent Output

Compare agent actions with what a trained employee would have done.

Validate Form Fill Assist


Confirm AI-generated entries don't conflict with validation or duplicate-detection rules.

Test Scheduling Scenarios

Use real operational constraints wherever possible.

Review AI Credit Consumption

Understand likely costs before production adoption increases.

Verify Security Controls

Check access behaviors carefully across environments.


What Should You Communicate Internally?


Many of these updates don't require formal training.



They require awareness.

Tell users:


  • When information comes from Copilot

  • When schedules may be adjusted automatically

  • When records may be AI-generated

  • When AI credit usage affects budgets


A short internal communication today can prevent confusion later.



Practical Recommendation


Don't read all the release notes.

Read the release notes that apply to your workloads.

Don't wait for production.


Test in sandbox.


And most importantly:

Treat every new agent capability as both a productivity feature and a governance decision.


The organizations that do that will benefit from these updates.

The organizations that don't will eventually discover them through support tickets.



Need Help Preparing for Release Wave 2?


If you're unsure which Dynamics 365 updates actually affect your environment, a release-readiness review can help identify:


  • Relevant feature changes

  • Potential risks

  • Configuration impacts

  • Recommended testing priorities


Before new functionality reaches production.

bottom of page