Business Central release wave 1/2026: From an ERP that documents to an ERP that collaborates
Share:
Most small and medium-sized enterprises have long since gone digital. Accounting, purchasing, warehousing, and reporting run on ERP systems; processes are documented, and data is available. Yet day-to-day operations in finance, procurement, master data management, and analysis remain surprisingly labour-intensive. Much depends on specific individuals, many actions are reactive, and the information residing in the system does not automatically lead to better decisions or more stable workflows.
This is precisely where Business Central release wave 1/2026 comes in. With the latest release of Microsoft Dynamics 365 Business Central, Microsoft is enhancing specific features while simultaneously driving a broader evolution: the wave illustrates how ERP systems are set to transform in the future through AI, Copilot, and autonomous agents. The following points provide a concise overview of the key new features in version 28 (BC28).
Key developments in Business Central release wave 1/2026
- Expansion of Copilot and AI agents
- Payables Agent for accounts payable
- Expense Agent for expense and travel cost processes
- New quality management in Business Central
- Enhancements for Shopify and commerce processes
- Improved analytics capabilities and item insights
- Enhanced governance and control for AI-driven processes
For managing directors and IT decision-makers, however, what ultimately matters less is any single feature than how the role of the ERP system itself is changing. This is precisely where the true significance of the ‘wave’ lies: Microsoft is gradually evolving Business Central from a system that merely documents data into an active work platform. This makes the wave strategically more significant than many updates seen in recent years.
The most striking change in this release wave is not a single feature, but a new approach to business data. In its official overview, Microsoft explicitly describes Business Central as a step towards an AI-powered ERP designed to guide companies from the manual execution of tasks to more autonomous, insight-driven processes. The system is intended to increasingly act as an active partner in day-to-day operations.
What sounds technical is, at its core, a shift in the division of labour. Until now, ERP systems followed a simple logic: the system stores data, while humans analyse, decide, and execute. In the future, the software will assume responsibility at an earlier stage in this chain. It will evaluate data, generate summaries and alerts, support decision-making, and—within clearly defined parameters—carry out initial operational steps itself.
The trend is thus moving away from systems that merely document human actions towards systems that analyse, provide support, and increasingly participate in execution. This shift is likely to shape the next generation of ERP projects more profoundly than any individual module function.
Copilot and AI agents are transforming the role of the ERP system
If you want to understand what Microsoft is trying to achieve, it is best to look at the expansion of Copilot and AI agents. This is where the key difference underpinning the entire wave lies.
Copilot remains the supporting element. It summarises content, provides insights, and makes information more readily accessible. AI agents go one step further. They carry out tasks and execute processes within defined rules, making their actions visible, controllable, and verifiable. In release wave 1/2026, Microsoft is expanding precisely this layer: agents are expected to handle more complete end-to-end processes, such as preparing an incoming invoice from an email, and work directly on pages within the system rather than in the background.
The Payables Agent effectively illustrates this principle. It monitors a designated mailbox, interprets incoming supplier invoices, and prepares the documents for further processing. A crucial detail here reveals the underlying logic of the wave: the agent creates drafts for approval rather than posting entries autonomously. The final decision remains with a human. The Expense Agent—which has also undergone further development—applies the same approach to expense and travel cost management.
This reveals a pattern running through the entire wave: Business Central is evolving from a data entry system into a platform that actively prepares operational tasks while deliberately leaving responsibility for them with the user. It is precisely this interplay—preparation by the agent and approval by the human—that serves as the common thread underpinning the wave’s control and governance functions.
How the transformation is manifesting in finance, supply chain, and commerce
It is noteworthy that this strategy does not stop at the Copilot level. It extends into the core ERP processes, and that is precisely where tangible benefits for medium-sized enterprises emerge.
In the finance sector, Microsoft is expanding the system beyond AI support for payables and expenses by introducing tangible new features. Self-billed invoices allow for the generation of invoices on behalf of the supplier, thereby increasing accuracy and processing speed while maintaining cleaner data. Both features—self-billing and standardised withholding tax logic for suppliers—will reach general availability in June 2026. Withholding tax thus becomes part of the standard functionality, eliminating the need for separate third-party solutions or custom development. The underlying strategy is clear: fewer manual steps, greater standardisation, and closer integration of efficiency, compliance, and control.
In the areas of supply chain and manufacturing, too, Microsoft is investing in more robust control mechanisms rather than mere speed. It is therefore significant that subcontracting and quality management are explicitly highlighted as key focus areas in the official overview. With its native quality management capabilities, Microsoft establishes quality inspections as an integral part of procurement, production, assembly, and warehousing processes—incorporating inspection rules with pass/fail criteria, the blocking of non-conforming batches, and quarantine and follow-up logic. In Microsoft’s framework, therefore, automation equates to reproducible quality and controllable operational decisions.
At first glance, the Shopify extensions might seem like a matter of fine-tuning. However, if product variants, attributes, images, and collections can be managed more consistently across systems, the effort associated with broken workflows, corrections, and manual reconciliation decreases. From a strategic perspective, this implies something more significant: Microsoft is increasingly viewing commerce and ERP as an integrated process landscape rather than as separate worlds.
Finally, the nature of analysis is also shifting. With features such as ‘item insights’ and AI-powered KPI summaries, the system is evolving from a mere reporting tool into a platform that interprets data within a specific context. After all, the value of ERP data lies not in its availability, but in its utility for making concrete decisions. This is precisely the threshold where this trend is clearly taking hold.
Why governance is becoming a prerequisite for AI in ERP
One point deserves special attention because it is equally central for decision-makers and consultants. Microsoft deliberately avoids a ‘black box’ approach; automation and control are intended to go hand in hand.
This is evident in the accompanying control mechanisms. A central task pane consolidates the activities of all agents in one place. Content generated by agents can be reviewed directly on the relevant pages. Indicators identifying the entities responsible for generation or modification provide transparency regarding where AI interacts with the data. Furthermore, a function to instantly halt all active tasks ensures the company retains full control at all times. At the same time, Microsoft emphasises that these enhancements for finance are designed to boost productivity and efficiency while maintaining control and compliance.
Microsoft thus views agent-driven processes as a controllable division of labour between humans and systems. Control deliberately remains with the company. This is a crucial factor for acceptance, particularly among small and medium-sized enterprises, as AI in ERP systems only becomes viable when it remains transparent and comprehensible. Consequently, governance shifts from a peripheral issue to the heart of every implementation.
The real challenge: organisational maturity rather than technology
As compelling as the technical direction may be, the real hurdle lies elsewhere: in the organisational maturity with which a company utilises these functions. Once agents take on or prepare tasks, questions arise that cannot be answered with a single click.
- What tasks can agents actually take on?
- Which decisions remain with humans?
- Is the data foundation robust enough to ensure reliable results?
- Are the relevant processes sufficiently standardised?
- Where can the benefits be measured, and what governance is required for this?
These considerations do not stem from a single Microsoft announcement; rather, they follow logically from the way Microsoft integrates Copilot, agents, control, and process automation. Autonomous or semi-autonomous workflows function reliably only when data is consistent, responsibilities are clearly defined, and processes are sufficiently stable. The crucial question, therefore, is not whether a company can activate these new features, but whether its organisation is prepared to use them productively and under proper control.
What is changing for ERP projects and consultancy
This shift also changes the logic of future ERP projects—a point relevant to both companies and consultants alike. After all, the focus has previously been primarily on traditional topics:
- Model business processes,
- define permissions,
- migrate master data,
- configure posting logic and workflows.
This work remains necessary, but it is no longer sufficient.
A new dimension is also being added. Going forward, the project will involve designing agent behaviour, defining guardrails and the scope for decision-making, establishing review mechanisms, quantifying the benefits of AI scenarios, and redefining the division of roles between humans and the system. When an ERP system moves beyond merely mapping processes to actively participating in them, the implementation becomes more demanding from both a functional and an organisational perspective.
For partners, consultants, and project managers, this wave is therefore also relevant from a methodological standpoint. The value of their work increasingly lies in providing sound answers to these new questions, rather than simply configuring functions.
Business Central release wave 1/2026 is not a standard feature release. Microsoft is delivering new capabilities for finance, supply chain, manufacturing, commerce, and analytics; some of these—such as self-billed invoices and withholding tax logic—will become available gradually over the course of the wave, leading up to September 2026. Microsoft itself notes that planned content and timelines remain subject to change prior to final release. Consequently, it would be premature to assess the feature set for production readiness at this stage.
The true significance of this wave runs deeper, however. With Copilot and AI agents, Business Central is gradually evolving from an ERP system that merely documents processes into an active work platform that supports—and in some cases executes—them. This creates genuine potential for efficiency gains for companies. At the same time, however, demands regarding data quality, process standardisation, governance, and project methodology are increasing.
For small and medium-sized enterprises, the crucial question is therefore no longer whether AI will become part of ERP systems. Instead, it is how a company will shape the collaboration between humans and systems in the future. Those who address this question early and deliberately will derive far more from this wave of innovation than mere functional gains.
Do you have any questions? We would be happy to discuss them with you. You can contact us directly via our contact form.
Back to the blog
