The artificial intelligence capabilities of Microsoft Dynamics 365 Business Central can help simplify an increasing number of everyday business tasks. Copilot can assist users with finding data, navigating the system, preparing documents and conducting analyses, while AI agents may eventually be able to support entire business processes with a greater degree of autonomy. In this article, we explore the opportunities these solutions offer and the key considerations businesses should address before implementing them.

 

In our previous blog post, we presented the most important financial, reporting, logistics and quality management enhancements introduced in the 2026 release wave 1 of Microsoft Dynamics 365 Business Central, also known as version 28. However, BC28 does not only bring changes to traditional business processes: Copilot and AI capabilities are also playing an increasingly prominent role within the system.

These developments are designed to help users find the information they need more quickly, spend less time on repetitive administrative tasks, and use natural-language instructions to retrieve information, prepare documents and create analyses.

 

Copilot and AI agents: what is the difference?

Business Central’s AI capabilities can be divided into two main categories.

Copilot primarily assists users in carrying out their work. It can search for information, generate suggestions, analyse data, produce text and help users navigate the system. The user remains actively involved throughout the process, reviews the results and typically makes the final decision.

Agents, or AI agents, are being developed to operate with a greater degree of autonomy. They may monitor specific business processes, communicate with users or business partners, and prepare the necessary Business Central records. However, these solutions are currently at varying levels of maturity, with several agent capabilities still in development or experimental stages.

The Copilot and AI developments in BC28 primarily cover the following areas:

• searching business data, navigating the system and assisting users with Business Central;
• automatically completing fields based on historical business data and information available online;
• generating marketing copy from information stored in the item master data;
• intelligently processing electronic documents and incoming invoices;
• reconciling bank accounts and matching transactions;
• preparing sales documents using natural-language instructions;
• generating record summaries and pivot-style analyses spanning multiple tables;
• using sales, purchasing and custom business agents.

The actual availability of these features may depend on the Business Central deployment model, region, system language and administrator settings. The broadest native support for Copilot and AI agents is available in cloud-based SaaS environments.

 

An AI-powered business assistant

Chat with Copilot can be used as a natural-language business assistant within Business Central. Users do not necessarily need to know the exact structure of the system’s menus or the names of individual pages. Instead, they can describe the information or functionality they need in everyday language.

The artificial intelligence capabilities of Microsoft Dynamics 365 Business Central can help simplify an increasing number of everyday business tasks.

The assistant can do more than search for data. It can also explain Business Central concepts and fields. For instance, it can show users where to find the chart of accounts and then navigate them directly to the relevant page.

When generating responses, Copilot takes the individual user’s permissions into account. It can therefore only work with business information that the employee is already authorised to access within Business Central.

Automatic field completion suggestions

AutoFill with Copilot can suggest values for editable fields that are still blank on Business Central cards. These suggestions may be based on historical Business Central records, business data available to the user and, where separately enabled, information found online. On a vendor card, for example, Copilot may suggest address and contact details. When preparing purchase orders, it may recognise frequently used cost centre and project codes based on previous documents.

When generating suggestions from data stored in Business Central, the system may consider values that have been used most frequently or most recently. This can reduce the time spent entering data, particularly when creating recurring documents with a similar structure.

Copilot highlights the suggested fields, and users can accept or reject them individually or in bulk. The information is not saved automatically, ensuring that final control remains with the user.

Online search functionality can be enabled or disabled separately. Before using this feature, businesses should therefore assess, from both data protection and compliance perspectives, what business data and contextual information may be transmitted to the AI service.

AI-generated marketing copy from item master data

Copilot can generate marketing copy directly from the information and attributes stored on an item card.

Users can specify the desired tone of voice, choosing from styles such as formal, casual, inspirational, upbeat or creative. They can also define the preferred format and instruct the system to emphasise innovation, sustainability, performance, elegance, reliability or other product characteristics.

Users can control which of the attributes assigned to an item Copilot should use when generating the text. Information that is not relevant to the specific communication objective can therefore be excluded.

This capability can be particularly useful when preparing content for online stores, catalogues and product pages. It can quickly generate an editable first draft from existing master data, which users can then adapt to the company’s communication style and brand voice.

Suggestions for substitute items

Based on product data, Copilot can automatically suggest substitute items for a selected product, providing a quick starting point for identifying suitable alternatives. This can be particularly useful in situations where an item is out of stock and potential replacements need to be identified quickly.

The Matching setting allows users to determine how strict the matching process should be, with options ranging from lenient to balanced or precise. The View option also lets users decide whether they want to see only the matches or additional information such as the match assessment and the keywords used.

The artificial intelligence capabilities of Microsoft Dynamics 365 Business Central can help simplify an increasing number of everyday business tasks.

Intelligent document and invoice processing

This Copilot capability, which is currently still under development, is designed to support the matching of incoming electronic invoice lines with purchase order lines.

The AI attempts to associate invoice lines with the appropriate order lines based on factors such as the description, unit of measure, amount and invoiceable quantity. The solution may also be able to handle more complex relationships rather than being limited to one-to-one matching.

This means that a single invoice line could potentially be matched with several purchase order lines or related entries. The feature may therefore reduce the time spent on manual reconciliation and data entry, particularly when processing large volumes of incoming documents.

In its current form, the solution is limited to SaaS environments, is available in English and requires a properly configured electronic document management environment as well as a connection to an external service provider. It is therefore particularly important to verify the technical requirements and test the functionality before implementation.

AI-assisted bank account reconciliation

Copilot builds on Business Central’s standard automatic matching functionality. Once the conventional automatic reconciliation process has been completed, the AI can examine the remaining unmatched bank transactions and ledger entries.

When searching for matches, it may take into account the amount, date, payment reference, transaction description and related business context. A single bank transfer may even be matched with several separate customer or vendor invoices.

Where no direct match is found, Copilot may suggest an appropriate general ledger account based on the transaction description. For example, it may recognise a bank charge or insurance premium and recommend the account to which the transaction should be posted.

Copilot can also help identify missing postings, duplicate entries and exchange-rate differences. However, final matching and posting decisions remain subject to user approval.

Creating sales documents using natural language

Copilot can also assist users in preparing the lines of sales quotes, orders and invoices.

Users can provide natural-language instructions to add specific products and quantities, copy lines from an earlier document, or repeat the most recent similar order. This can be particularly helpful when a document contains numerous recurring items or lines that are already known from previous transactions.

The feature also supports the processing of CSV files, allowing document lines to be imported from an external file quickly and with only a few clicks.

The system can search for products not only by item number but also by description. Users therefore do not necessarily need to know internal product identifiers from memory.

Based on the instruction, Copilot prepares the document lines, which the user can review, edit and approve before the document is finalised.

Automatic record summaries and business insights

The Summarize Records feature presents the most important information associated with an open record in a concise, easy-to-review summary.

The artificial intelligence capabilities of Microsoft Dynamics 365 Business Central can help simplify an increasing number of everyday business tasks.

The summary can appear in a separate side panel, without requiring the user to write a new prompt each time. This allows users to gain a quick overview of a complex record without reviewing every field and related document individually. The feature works exclusively with data that the user is authorised to access.

Pivot-style reports using natural-language instructions

The Analyze Data in Lists feature builds on Business Central’s analysis mode. It allows users to describe the report or analysis they want to create using natural language.

Based on the request, Copilot can select the necessary fields and configure groupings, aggregations, filters and sorting. The result may be an editable analysis view similar to a traditional pivot table.

The feature may also support the combination of data from multiple tables. For example, users could request a list of vendors ranked by balance or an analysis displaying related document numbers alongside those balances.

The completed view can be refined further and saved at user level for future use. For more complex analyses involving multiple tables, instructions written in English may currently produce more accurate results. Businesses should therefore test the queries that are most important to their operations before deploying the feature in a live environment.

 

AI agents in sales and purchasing processes

Agents represent one of the fastest-developing areas of Business Central. Some of the capabilities currently being presented are still in development or experimental stages. At this point, they primarily illustrate the expected direction of development and the potential for future automation.

Sales Order Agent

The Sales Order Agent is designed to process order requests received in a dedicated email inbox and prepare the related sales document.

Where a request is unclear, the agent may send a follow-up message to the customer asking for additional information. Based on the response, it can update the draft sales quote or order, which the user can review before finalisation.

The feature is currently still under development, and its operation in a Hungarian-language environment requires separate testing.

Payables Agent

The Payables Agent is intended to support the processing of invoices received from vendors by email.

Under the proposed workflow, the agent may use the documents to prepare a draft purchase invoice, identify the vendor and invoice lines, and then submit the document for user review.

The feature cannot yet be regarded as a finished solution ready for general deployment. Nevertheless, it provides a clear indication of how vendor invoice processing may become increasingly automated in the future.

Custom business agents

Business Central also offers the possibility to design and test custom business agents. An agent’s task can be defined using natural-language instructions, while its activities can be controlled through permissions.

In the future, custom agents may be used for tasks such as checking data, preparing records, creating document drafts and supporting repetitive administrative activities.

However, the current solution remains an experimental tool. It is primarily suitable for testing manually initiated tasks and should not yet be considered a comprehensive, autonomous workflow engine.

Before using agents in live business processes, organisations should carefully assess the maturity of the individual features, language support, permissions, data management requirements and the points at which user approval is required.

 

What does this mean for Business Central users?

Copilot and AI capabilities are not limited to a single business function. They can support a wide range of everyday tasks, from data search and system navigation to financial reconciliation, the preparation of sales documents and business analysis.

Copilot can primarily make users’ work faster and simpler, while decision-making and approval remain under human control. Over time, agents may enable a higher degree of automation, although their current maturity and availability vary from one feature to another.

Successfully introducing these capabilities therefore involves more than simply enabling the AI functions. Businesses need to assess their technical environment, access permissions, data protection requirements and language support. They must also determine which capabilities can create genuine value within their specific business processes.

DynaGo’s experts can help identify which of the Copilot and AI capabilities available in Business Central are best suited to your organisation’s operations. We provide support in assessing the technical and business requirements, testing the relevant features, and planning a secure and controlled implementation.

Contact us to ensure that AI becomes more than just another new technology—and delivers genuine, measurable operational benefits for your business.