When Christian Klein, SAP CEO, recently wrote his column “AI isn’t killing SaaS—it’s making it indispensable” in PLM & ERP News, he highlighted an issue many organizations are currently grappling with: where is the real value created in the shift to AI? Because as AI capabilities accelerate, a misconception also grows: that the intelligence itself – the models – is central. In practice, it’s often the opposite.
“What Christian Klein highlights is an important perspective. AI itself is becoming more and more powerful, but without structure, context, and business logic, it quickly becomes quite limited. It is within ERP systems that it finds a context in processes and data that creates value and direction,” says Johan Berg, Business Development Manager SAP at Implema.
From System Support to Decision Engines
Traditionally, business systems – ERP, CRM, and other business-critical platforms – have functioned as systems of record. They have documented what has happened, created structure and traceability in business processes, and generated data for analysis and follow-up. With AI, this role is fundamentally changing.
“We are seeing how systems are moving from being merely documenting to becoming proactive, suggesting actions, and identifying deviations. It’s no longer just about follow-up, but about predicting, proposing, and in some cases, automating decisions in real-time,” says Johan.
Examples already exist in operations:
- Quotation processes that are automated, dramatically shortening lead times
- Planning that is continuously adjusted based on changing conditions
- Administrative work replaced by AI-driven workflows
This is where Klein’s reasoning takes on a very concrete meaning: AI needs a place to operate. And that place is the ERP systems.
Data Quality Becomes Business-Critical – For Real
One consequence of this development is that issues previously considered “IT issues” are rapidly moving up to management level.
“When AI starts making or suggesting decisions, it’s no longer enough for data to be ‘good enough.’ It must be accurate, consistent, and owned. Otherwise, you’re just scaling up inaccuracies.”
This places higher demands on:
- Data governance and ownership
- Common definitions across the organization
- Standardized processes
Here, SAP’s focus on Clean Core and standardization also becomes central, not as a “technical ideal,” but as a prerequisite for leveraging AI in practice.
From Fragmentation to Context
Another important implication is how organizations view their system landscape. For a long time, “best of breed” has been a dominant strategy, meaning choosing the best systems within each area. But in an AI-driven reality, the consequences of fragmentation become clearer.
“AI requires context. When data and processes are scattered in silos, intelligence also becomes fragmented. Then the desired effect is lost.”
It is in this light that the hyped collaboration between SAP and Siemens should be understood, where business data and engineering data are linked together in a common context.
The System Landscape Determines the Outcome
Klein’s point can also be understood through how value has historically shifted in technological transitions. Initially, value is created in infrastructure and core technology. But over time, it moves upwards, to the application layer, where technology is translated into business benefits.
“Many organizations are currently focusing on AI as a technology. But what actually determines the outcome is their system landscape, i.e., how well data, processes, and systems are interconnected. Here, ERP systems are not a limitation, but an incredibly important enabler based on the processes and built-in integrations they possess,” says Johan.
AI Amplifies – It Does Not Replace
Overall, both SAP and experiences from implementations in the Nordics point in the same direction: AI is not a disconnected layer that can be “added on top.” It is an amplifier.
“If you have structure, clear processes, good data quality, and a well-thought-out system architecture, then AI can create enormous leverage. If you don’t, you risk merely automating inefficiency.”
A Strategic Question for Management
This also makes the AI question fundamentally strategic:
- What does our information model look like?
- Where are decisions made – and on what basis?
- How are our systems and processes interconnected?
“It is only when these questions are answered that AI can begin to create real value. And in that regard, ERP systems play a more central role than one might think,” concludes Johan Berg.
