AI and ERP – over the past few months, one assumption has seemed almost inevitable: if you want to stay ahead with artificial intelligence, sooner or later you will need a completely new ERP system. Preferably cloud-based. Preferably everything from scratch.
The current SAP Business One roadmap paints a different picture.
SAP Business One is here to stay. And around it, a new AI ecosystem is taking shape step by step.
That may be the most interesting message in the entire roadmap. SAP is not simply adding a few new AI features. When you look at the individual building blocks together, a fairly clear direction emerges.
And it starts in a surprisingly pragmatic way.
Step 1: Build the Foundation First, Then Add AI
Before an ERP system can answer questions, identify connections or even take over tasks at some point, it first needs the technical foundation to do so.
That is exactly where the roadmap begins.
SAP Business One remains the ERP system and therefore the home of core business processes and company data. New technologies are added via the SAP Business Technology Platform and AI services.
That may sound more technical than it really is.
Put simply:
The house stays where it is. But the technology inside it gets smarter.
For existing SAP Business One users, that is an important message. Future-proofing does not necessarily mean throwing out a proven ERP system and starting again from scratch.
Step 2: The ERP Starts to Understand Our Language
The next building block on the roadmap will be much more visible to users.
Today, anyone who needs information from an ERP system usually still has to know where to look: Which report? Which analysis? Which filter? And in some cases, even which SQL query?
In future, a simple question could often be enough:
“Which customers are ordering significantly less than last year?”
Or:
“Which open quotations currently need my attention?”
The roadmap clearly points towards natural language and context-aware AI assistance.
And this is where things become really interesting. Because this is not simply about adding another feature. It could change the way people interact with an ERP system.
The user no longer has to understand the logic of the system.
The system begins to understand the user’s question.
And this is where a single feature turns into a much bigger story.
Because AI can only provide a meaningful answer if it understands the right context.
And in most companies, that context is rarely stored in just one place.
Step 3: AI Needs Access to Business Knowledge
Customer data may sit in the ERP system. Documents may be stored in a DMS. Communication may live in Microsoft 365. Additional information can be found in CRM, production, service or specialised applications.
For Business AI to become genuinely useful, these different worlds need to work together in a meaningful way.
This is where a term appears in SAP’s roadmap that is worth remembering:
MCP – Model Context Protocol.
Don’t worry: if your first instinct was to close the browser window at this point, you can stay. 😉
The basic idea is far easier to understand than the name suggests.
MCP is intended to standardise how AI applications can access data, tools and functions in a controlled way.
We explain in more detail in our webinar why this could become particularly relevant for SAP Business One. For now, one idea is enough:
The better AI understands the business context, the more valuable it becomes.
And that is exactly why integration is suddenly taking on a completely new role.
Step 4: When Assistance Becomes Genuine Support
So far, we have mainly been talking about AI finding information, putting it into context and presenting it in an understandable way. The next stage goes further: systems could eventually do more than simply respond to questions. They could recognise on their own when something needs attention.
Take a simple example. An important customer suddenly orders significantly less than usual. At the same time, there are open service cases and a major quotation has been sitting in the system for an unusually long time. Today, someone would have to collect this information from different sources and connect the dots manually.
An AI-powered agent could eventually recognise this context automatically, bring the relevant information together and suggest a sensible next step – for example, preparing a sales task or highlighting that action may be required.
The important point is that this is not about AI suddenly “running the company”. The more interesting idea is that AI can actively support people in areas that still require a great deal of manual attention today.
That is why AI agents are such an interesting part of the long-term development of SAP Business One. They show where the journey could lead: away from software that simply waits for input and towards systems that recognise connections and support users proactively.
And Then an Old Topic Suddenly Becomes More Important Than Ever
The more AI becomes part of everyday business, the more important the quality of the information it can access becomes. Even the best AI cannot establish reliable connections if customer records are duplicated, information is inconsistent across systems or important details are simply missing.
That is why a topic that has long been regarded as a rather tedious IT housekeeping task is suddenly gaining strategic importance: data quality.
For traditional reporting, incomplete data has always been frustrating. For AI, it becomes a real problem. If a system is expected to prepare decisions, identify risks or evaluate connections, it is not enough to simply have “a lot of data”. That information needs to be up to date, consistent and, above all, connected.
This is also where the role of integration and data platforms begins to change. They are no longer just there to make systems communicate technically. They create relationships between information – and that context will ultimately determine whether AI merely finds data or actually understands the business behind it.
For us, this is also where C.ONE comes into the picture: not as an “AI solution” in itself, but as a connecting layer between data, applications and processes. The stronger this foundation is, the more value future AI functions can deliver on top of it.
Perhaps that is one of the most important lessons from the entire roadmap: preparing for Business AI does not begin with choosing an AI model. It starts much earlier – with the question of how well your data and systems already work together today.
So What Does the Roadmap Really Tell Us?
Perhaps less:
“SAP Business One is getting a bit of AI now.”
And much more:
SAP is gradually building an architecture in which Business One, business data and artificial intelligence can work together.
That is a significant difference.
And especially for companies with existing SAP Business One environments, this makes the coming months particularly interesting.
Because the questions are now becoming very concrete:
Which of these capabilities are already available?
What is coming next?
What is really behind Ask AI and MCP?
Where does concrete planning end – and where does the vision begin?
And above all:
What does this development mean for companies that want to continue running SAP Business One on-premise?
These are exactly the questions we will take a closer look at in our upcoming webinar.
Learn More in Our Live Webinar on 18 August 2026 at 2:00 p.m.
SAP Business One 2026: The AI Roadmap Is Here – and What It Means for On-Premise
In around 25 minutes, we will take a closer look at the current roadmap, walk through the individual development stages and explain what is already planned, which technologies are behind it and which topics are still part of the longer-term vision.
No AI hype. Just a clear and practical look at what these developments could actually mean for day-to-day business.
→ Register now for the free webinar
