We need to talk about... Dynamics 365 Customer Service... Release Wave 1 for 2026

Today in my series on the changes we can expect to see as part of Release Wave 1 for 2026, Dynamics 365 Customer Service is next. I will take a look at what's actually involved in this release for our Customer Service implementations.

This wave is about giving the people managing and delivering that service better tools to do it: smarter quality control, safer AI deployment, more proactive insight into how customers are feeling, and forecasting that actually works.

Case Management Agent

If you've been following what Microsoft has been doing with the Case Management Agent the AI that can autonomously handle and resolve cases you'll know that the capability is genuinely impressive.

This feature answers that question directly. It gives administrators the ability to run full end-to-end simulations of case resolution flows before anything touches a real customer. You can simulate how the agent would handle cases from identifying customer intent, through drafting a response, to calling tools and completing actions using either synthetic data or a sample of your actual organisational records. And crucially, no emails are sent and no case data is modified. It's a safe sandbox for testing.

Each simulation run captures detailed output: the rules and instructions the agent followed, the responses it drafted, the tools it called, any issues detected, and an estimate of the credit cost. Everything is tagged and versioned, so you can compare runs side-by-side as you refine your configuration.

Improved Sentiment Analysis

Sentiment indicators on a case are designed to prevent exactly that scenario. This feature analyses the emotional tone across every interaction on a case email, chat, and voice weights them by recency and by channel, and surfaces a single, clear sentiment score directly on the case form and in the active cases grid. Positive, neutral, or negative: visible at a glance, before the agent has read a word.

It updates automatically whenever a new interaction arrives, so if a customer's mood shifts mid-case, that change is reflected immediately. Supervisors can configure which channels carry more weight for example, giving voice interactions a higher weighting than email so the scoring reflects how your organisation actually operates.

Shadow Mode in Cases

Shadow mode is Microsoft's answer to that. When you enable it for the Case Management Agent, the system runs in parallel with your human agents. It analyses every incoming case, identifies the customer intent, drafts a response, suggests field updates, and recommends actions exactly as it would if it were live. But nothing is sent. No case is modified. No customer sees anything.

Instead, all of that predicted output is captured in a dedicated shadow results view on the case record. Administrators and supervisors can review it at their leisure, compare the agent's recommendations to what the human agent actually did, identify gaps or mismatches, and build the evidence base they need to make an informed decision about moving to partial or full automation.

Qualify Evaluation Agent

In scenarios where agents forget to read a data protection disclosure, or fails to verify the customer's identity before discussing account details, or misses a mandatory product warning. In a standard scored evaluation, these failures might be partially offset by strong scores elsewhere. The agent passes overall. The compliance failure is buried.

Critical questions fix that. Supervisors can mark one or more questions within an evaluation criterion as critical and when a critical question is failed, the entire evaluation is automatically marked as failed, regardless of every other score. There's no offsetting, no partial credit, no ambiguity. If the non-negotiable wasn't met, the evaluation doesn't pass.

It integrates fully with the Quality Evaluation Agent framework, including versioning and simulations, so supervisors can test their criteria before publishing them for live use.

Screen Recording Capability

Microsoft has introduced enhanced screen recording capabilities in Dynamics 365 Customer Service, giving organisations greater visibility and control over how customer interactions are captured and reviewed.

At its core, the update strengthens governance and oversight. Role-based access ensures only authorised users can view or download recordings, supporting compliance while maintaining operational transparency. Supervisors can also play recordings directly within the Customer Service workspace, helping speed up investigations and reduce the need to switch between tools.

From a business perspective, this translates into faster issue resolution, reduced rework, and more effective coaching. By combining screen context with interaction data, organisations gain a clearer understanding of how work is actually performed—enabling more accurate quality reviews and targeted performance improvements.

Summary

Release Wave 1 for Dynamics 365 Customer Service is a massive step forward in building confidence in your AI tools, in your agents, and in your data.

The simulation and shadow mode features give organisations a structured, evidence-based path to AI adoption. The sentiment indicators and knowledge adherence evaluations give supervisors visibility they've never had before. Critical questions bring proper compliance rigour to quality management. And dynamic forecasting removes the guesswork from one of the most consequential decisions a contact centre makes every single day.

Hope you all enjoyed this post on the Dynamics 365 Customer Service Release Wave 1 for 2026. I will continue this series with some additional content specifically related to Release Wave 1 for 2026. Take Care, Speak Soon!