Microsoft Dynamics 365 Copilot: The Complete Guide for 2026

Quick Answer: What Is Microsoft Dynamics 365 Copilot?

Microsoft Dynamics 365 Copilot is Microsoft's AI layer for its CRM and ERP applications. It combines built-in Copilot assistance with AI agents that can analyze business data, summarize records, generate content, recommend actions and, in supported scenarios, execute multi-step business processes.

In 2026, the important distinction is that “Dynamics 365 Copilot” is no longer one feature or one product. Organizations may use:

• Built-in Copilot capabilities inside individual Dynamics 365 applications
• Dynamics 365 agents for specialized sales, finance, service and supply-chain workflows
• Microsoft 365 Copilot to work with business information alongside Outlook, Teams, Word, Excel and other Microsoft 365 applications
• Work IQ to provide Microsoft 365 Copilot and agents with permission-aware organizational context
• Microsoft Copilot Studio to create and extend custom agents
• Copilot Credits for certain metered AI and agent workloads

When evaluating Dynamics 365 AI, organizations need to look beyond the simple question, “Do we have Copilot?” The real question is which Copilot capabilities your users need, what data those capabilities can access, how they are licensed, and what controls should be in place before deployment.

Key Takeaways

• Microsoft is shifting Dynamics 365 from primarily assistive AI toward AI agents capable of completing increasingly complex business tasks.
• Copilot capabilities are available across Dynamics 365 Sales, Customer Service, Field Service, Finance, Supply Chain Management, Business Central, Customer Insights and other Dynamics applications.
• Some Copilot functionality is included with qualifying Dynamics 365 licenses, while agent usage may require Copilot Credits.
• Microsoft 365 Copilot and Dynamics 365 are becoming increasingly connected through Dataverse and Work IQ.
• Data quality, permissions, security and governance should be addressed before organizations deploy Copilot broadly.
• Microsoft is retiring the traditional twice-yearly Dynamics 365 release-plan model beginning in September 2026 and moving new roadmap information to its continuously updated AI at Work roadmap.

How Dynamics 365 Copilot Has Changed

When Microsoft introduced Dynamics 365 Copilot in 2023, much of the emphasis was on generative AI assistance. A salesperson could draft an email. A service representative could summarize a case. A user could ask for instructions on completing a task. Those capabilities remain useful, but the direction of Microsoft’s platform has changed significantly.

The emerging model is agentic AI.

Traditional automation: “When an invoice becomes overdue, send this predefined email.”

Copilot assistance: “Draft an email to this customer about the overdue invoice.”

Agentic AI: “Review overdue accounts, identify the accounts that need attention, evaluate the available information and take the appropriate collection actions within our approved rules.”

The third model has much greater potential, but it also requires stronger governance because the AI is doing more than generating text.

What Does Copilot Do Across Dynamics 365?

Microsoft now describes Dynamics 365 as combining Copilot experiences, built-in AI and intelligent agents across its CRM and ERP applications. The capabilities differ considerably by application.

Availability, licensing and release status vary by feature, region and Dynamics 365 license, so organizations should verify individual capabilities before designing a production workflow.

Copilot in Dynamics 365 Sales

Dynamics 365 Sales is one of the most developed areas of Microsoft’s business AI strategy. Built-in Copilot functionality can help sellers prepare for meetings, summarize information, work with CRM data and draft communications. Microsoft’s newer sales agents go further by handling parts of the sales process on the seller’s behalf.

Sales Qualification Agent

Microsoft’s Sales Qualification Agent can research leads and assess them against criteria configured by the organization. Microsoft currently provides two operating modes.

Research-only mode researches assigned leads, generates information about the lead and company, evaluates fit and can prepare an initial outreach email.

Research and engage mode goes further. The agent can research and engage leads, follow up and evaluate whether the lead meets handoff criteria before sending qualified opportunities to sellers.

For a sales organization with a large inbound lead volume, this changes where AI can create value. Instead of simply helping a salesperson write faster, AI can help determine which opportunities deserve the salesperson’s attention in the first place. That does not eliminate the need for salespeople. It makes lead-selection criteria, CRM data quality and agent handoff rules much more important.

Copilot in Dynamics 365 Customer Service

Customer Service is another natural use case because representatives often need to understand large amounts of information before responding to a customer. Copilot can help representatives summarize cases and conversations, find relevant information, draft contextual responses, explore business data using natural language, and work with agents available through Microsoft 365 Copilot.

Microsoft has also brought Microsoft 365 Copilot directly into supported Dynamics 365 applications. This is important because the difference between “Copilot inside Dynamics” and “Copilot outside Dynamics” is becoming less defined. A service representative may increasingly interact with Dynamics data, Microsoft 365 information and specialized agents through the same Copilot experience.

Copilot in Dynamics 365 Finance

Finance is one of the areas where organizations should evaluate Copilot based on business outcomes rather than novelty. Potential applications include reducing time spent researching accounts, summarizing financial information, assisting collections teams, explaining data and supporting reconciliation and financial processes.

The value proposition is clear: finance professionals frequently spend substantial time collecting information before they can make a decision. AI can reduce that information-gathering burden. The risk is equally clear. Financial workflows are consequential.

Organizations should avoid treating AI-generated recommendations as automatically correct simply because they are based on ERP data. Approval thresholds, exception handling, data permissions and human review should be designed around the financial importance of each task.

Copilot and Agents in Dynamics 365 Supply Chain Management

Supply Chain Management illustrates particularly well how Dynamics 365 is moving from Copilot assistance toward operational agents. Microsoft’s Procurement Agent is designed to assist procurement teams with supplier changes and purchase-order workflows.

Current capabilities include supplier communications and impact analysis. For example, the agent can help identify purchase orders requiring supplier follow-up and generate communication to vendors. It can also process supported supplier messages and surface proposed changes for review.

Impact analysis can evaluate how supplier changes could affect inventory, production schedules and customer deliveries. This is substantially different from simply asking a chatbot a supply-chain question. It is AI becoming part of the procurement workflow itself.

Some Procurement Agent impact-analysis features remain in preview, so organizations should distinguish carefully between generally available functionality and preview functionality before making production decisions.

Copilot in Dynamics 365 Business Central

Business Central brings Microsoft’s AI strategy to small and midsize organizations that may not require the complexity of enterprise Dynamics 365 Finance and Supply Chain Management. Copilot and agent capabilities can help users work with ERP data through natural language, create and enhance business content, support operational tasks and financial processes, assist with sales and purchasing workflows, and extend Business Central with additional AI features.

For smaller organizations, this can be particularly valuable because the same employee may perform several operational roles. However, smaller companies should not assume that an AI agent requires less governance simply because the organization is smaller. An agent with access to financial, customer or purchasing records should still have carefully defined permissions and responsibilities.

Copilot in Dynamics 365 Field Service

Copilot in Field Service is designed to reduce the information burden on both managers and frontline workers. Field-service employees frequently need information from work orders, equipment histories, customer records and previous service interactions while working away from a traditional office. AI can help surface that information and reduce the amount of time spent searching through records.

Copilot in Dynamics 365 Customer Insights

Customer Insights applies AI to another data-intensive problem: understanding customers and deciding how to engage them. Depending on the Dynamics 365 Customer Insights application and capabilities being used, Copilot and agents can help marketers and customer-experience teams analyze data, develop segments, create journeys and improve customer engagement.

Here again, underlying data quality matters. An AI system cannot compensate for customer records that are duplicated, outdated, incorrectly classified or fragmented across poorly integrated systems.

What Is Work IQ?

Work IQ is the intelligence layer behind Microsoft 365 Copilot and Microsoft agents. It should not be thought of as another CRM or ERP application. Instead, Work IQ gives Copilot and agents organizational context by drawing on information and relationships across Microsoft 365 and connected business systems while respecting the user’s permissions.

For Dynamics 365 organizations, the important development is the connection between Microsoft 365 information and Dataverse business data. That can allow an employee to work with information from CRM or other business applications without repeatedly switching between separate interfaces.

• A Dynamics 365 customer record
• Outlook email
• Teams conversations
• Documents
• Meetings
• Other business information

Microsoft announced that the Work IQ APIs became generally available on June 16, 2026.

Dynamics 365 Copilot vs. Microsoft 365 Copilot

Increasingly, the more useful question is not which platform “owns” Copilot. It is what information the employee needs and where that work should take place.

How Much Does Microsoft Dynamics 365 Copilot Cost?

There is no single “Dynamics 365 Copilot price.” Organizations generally need to consider three cost layers.

1. Dynamics 365 Licensing

Many built-in Copilot capabilities are included with qualifying Dynamics 365 licenses. The specific entitlement depends on the application and license level. More advanced Dynamics 365 agents may require Copilot Credits.

2. Microsoft 365 Copilot

As of September 2026, Microsoft lists Microsoft 365 Copilot for enterprise customers at $30 per user per month when paid yearly, in addition to a qualifying Microsoft 365 subscription. Microsoft also offers Microsoft 365 Copilot Business for organizations with up to 300 users. Pricing and promotional offers can change, so companies should confirm current pricing when preparing a licensing estimate rather than relying on an older published figure.

3. Copilot Credits

Copilot Credits are Microsoft’s consumption unit for supported AI and agent workloads. Credit consumption depends on what an agent does and the complexity of the operation.

Microsoft currently includes 1,000 Copilot Credits per user per month with Dynamics 365 Sales Premium, Customer Service Premium, Finance Premium, and Supply Chain Management Premium. These credits are pooled at the tenant level.

Additional capacity can be purchased using Microsoft’s available pay-as-you-go or pre-purchase options. Since agent usage can vary substantially between organizations, a realistic Copilot budget should model actual workflows rather than multiplying users by a single AI price.

Why Data Readiness Matters More With AI Agents

Copilot does not magically repair poor business data. If customer records are duplicated, permissions are overly broad or key fields are inconsistent, AI can make those existing problems more visible and more consequential. Before a broad Copilot rollout, organizations should examine:

1. Data quality: Are the records Copilot will use accurate, current and consistently structured?

2. Permissions: Can users access information they do not actually need for their roles?

3. Sensitive information: Are confidential records properly classified and protected?

4. Agent permissions: What can each agent read, create, update or communicate?

5. Human approval: Which actions can occur automatically, and which should require review?

6. Monitoring: Can administrators understand what agents are doing and how much AI capacity they are consuming?

This becomes increasingly important as organizations move from AI that recommends actions to AI that performs them.

IES Perspective: Treat Copilot as a Business-Process Project

The technology is only one part of a Dynamics 365 Copilot implementation. The first question we recommend organizations answer is not “Which AI features should we enable?” It is: Which business process are we trying to improve?

Start with a measurable problem, such as slow lead qualification, repetitive supplier follow-up, excessive case-research time or inefficient collections work. Then determine what information the process requires, whether that information is reliable, who should have access, which steps can safely be automated, where human approval remains necessary, and how success will be measured.

That creates a much stronger implementation plan than enabling AI across an organization and hoping employees discover useful applications.

IES is a Microsoft Solutions Partner for Business Applications with additional Microsoft partner designations in Data & AI and Data & App Innovation on Azure. IES states that its team has worked on more than 600 Microsoft Dynamics projects over more than two decades. That experience supports a practical principle that applies to Copilot as much as it does to ERP and CRM implementations: technology produces the best results when it is aligned with a clearly defined business process.

What Changed in Dynamics 365 AI in 2026?

One of the most important 2026 developments is bigger than an individual Copilot feature. Microsoft is changing how Dynamics 365 itself evolves. Historically, Dynamics 365 and Power Platform customers planned around two large release waves each year.

Beginning in September 2026, Microsoft is retiring the twice-yearly release-wave disclosure model for new Dynamics 365, Power Platform, and Dataverse roadmap content. New capabilities will instead appear through Microsoft’s AI at Work roadmap, where features can move through stages such as development, rollout and launch continuously. Microsoft has also said there will not be a September 2026 Release Wave 2 announcement or release plan.

For Dynamics administrators, this means release planning should become more continuous as well. Rather than checking Microsoft’s roadmap only around major release-wave announcements, organizations should establish a regular review process for new Copilot, agent and Dynamics capabilities.

Is Dynamics 365 Copilot Worth It?

For some organizations, yes. But not because every employee needs every Copilot capability. AI is often most useful for routine business processes that involve repetitive information gathering, data analysis, communication, record handling, system updates, or other predictable tasks that require multiple steps to complete.

Copilot is less compelling when the underlying process is poorly defined, the data is unreliable or the organization cannot identify what outcome it expects AI to improve. In those cases, deploying more AI can automate confusion rather than eliminate it.

A Practical Dynamics 365 Copilot Implementation Checklist

1. Identify the business case. Choose a specific workflow and define the current cost, delay or productivity issue.

2. Document the data involved. Determine which Dynamics 365, Dataverse, Microsoft 365 and third-party information the AI needs.

3. Audit permissions. Verify that users and agents have access to what they need, but not significantly more.

4. Confirm licensing. Identify which Copilot capabilities are already included and which require Microsoft 365 Copilot or Copilot Credits.

5. Define agent boundaries. Specify what the agent may do automatically and where human approval is required.

6. Pilot with a limited group. Use a controlled set of users and workflows before organization-wide deployment.

7. Measure business outcomes. Track metrics such as time saved, response time, conversion, error rate or workload reduction instead of simply counting Copilot usage.

8. Review and expand. Scale successful use cases and modify or discontinue ones that fail to create measurable value.

Getting Started With Dynamics 365 Copilot

Dynamics 365 Copilot is quickly becoming less about adding a chatbot to CRM or ERP and more about changing how people and AI work together across business processes. This makes having a clear implementation strategy more important than simply turning on features.

IES helps organizations evaluate, implement, customize and support Microsoft Dynamics 365 environments. A Copilot engagement can include identifying appropriate use cases, reviewing Dynamics architecture and data readiness, evaluating licensing, configuring workflows and agents, and planning governance before deployment.

If your organization is evaluating Microsoft Dynamics 365 Copilot, start by identifying the business process you want AI to improve. From there, the right combination of Dynamics 365, Microsoft 365 Copilot and AI agents becomes much easier to determine.

Microsoft Copilot Frequently Asked Questions